From charlie at begeistert.org  Mon Mar  2 20:04:46 2009
From: charlie at begeistert.org (Charlie Clark)
Date: Mon, 2 Mar 2009 20:04:46 +0100
Subject: [Image-SIG] playing with CMYK images
In-Reply-To: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
References: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
Message-ID: <DDBEBD8A-20CB-4FEF-9EBC-A5D0174BF113@begeistert.org>


Am 27.02.2009 um 19:09 schrieb Ask Holme:

> Hey list
>
> i have just tried using PIL to crop and resize CMYK images (with  
> CMYK as end result, no color mode conversion). I must say i was  
> surprised how awfull the result look. Even though PIL is not  
> supposed to touch the color mode, colors is totally fucked up in the  
> modified image.
>
> Is that because i'm doing something wrong or is the only solution  
> simply to switch to imagemagick if i want better CMYK support

wow! As someone who normally appreciates informality on mailing lists  
and relishes in the triumph of the vernacular I found this a rude  
first post.

Can you please provide some details as to what's gone so horribly  
wrong. I would be surprised if the problem is actually with PIL per se  
and not with one of the underlying C libraries.

Charlie
--
Charlie Clark
Helmholtzstr. 20
D?sseldorf
D- 40215
Tel: +49-211-938-5360
GSM: +49-178-782-6226




From tim at timhatch.com  Mon Mar  2 20:32:32 2009
From: tim at timhatch.com (Tim Hatch)
Date: Mon, 2 Mar 2009 11:32:32 -0800
Subject: [Image-SIG] playing with CMYK images
In-Reply-To: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
References: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
Message-ID: <70DCC813-8501-4B29-A2EF-BD8190465A2A@timhatch.com>

> i have just tried using PIL to crop and resize CMYK images (with  
> CMYK as end result, no color mode conversion). I must say i was  
> surprised how awfull the result look. Even though PIL is not  
> supposed to touch the color mode, colors is totally fucked up in the  
> modified image.

Can you provide an example of your image and your cropping script?   
The first thing that comes to mind is that there is a color profile  
attached to the original, and it's being stripped.

Tim


From d_berthelot at yahoo.com  Mon Mar  2 22:14:57 2009
From: d_berthelot at yahoo.com (David Berthelot)
Date: Mon, 2 Mar 2009 13:14:57 -0800 (PST)
Subject: [Image-SIG] playing with CMYK images
References: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
	<70DCC813-8501-4B29-A2EF-BD8190465A2A@timhatch.com>
Message-ID: <430166.97363.qm@web52108.mail.re2.yahoo.com>


I ran into a similar problems with CMYK and I think it's because libjpeg doesn't interpret the CMYK data correctly (possibly stripping a color profile). This is easily verified using libjpeg standalone tools (like jpegtran that you can compile when compiling libjpeg).

On the other hand, using ImageMagick python bindings (PythonMagick) worked fine (I assume ImageMagick doesn't rely on libjpeg). However it is my understanding that PythonMagick is unfortunately not fully fonctional and barely documented.



----- Original Message ----
From: Tim Hatch <tim at timhatch.com>
To: image-sig <image-sig at python.org>
Sent: Monday, March 2, 2009 11:32:32 AM
Subject: Re: [Image-SIG] playing with CMYK images

> i have just tried using PIL to crop and resize CMYK images (with CMYK as end result, no color mode conversion). I must say i was surprised how awfull the result look. Even though PIL is not supposed to touch the color mode, colors is totally fucked up in the modified image.

Can you provide an example of your image and your cropping script?  The first thing that comes to mind is that there is a color profile attached to the original, and it's being stripped.

Tim

_______________________________________________
Image-SIG maillist  -  Image-SIG at python.org
http://mail.python.org/mailman/listinfo/image-sig



      

From jcupitt at gmail.com  Mon Mar  2 23:58:04 2009
From: jcupitt at gmail.com (jcupitt at gmail.com)
Date: Mon, 2 Mar 2009 22:58:04 +0000
Subject: [Image-SIG] playing with CMYK images
In-Reply-To: <430166.97363.qm@web52108.mail.re2.yahoo.com>
References: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
	<70DCC813-8501-4B29-A2EF-BD8190465A2A@timhatch.com>
	<430166.97363.qm@web52108.mail.re2.yahoo.com>
Message-ID: <522c6460903021458y10929c48sba7f89cf766a6d88@mail.gmail.com>

2009/3/2 David Berthelot <d_berthelot at yahoo.com>:
> I ran into a similar problems with CMYK and I think it's because libjpeg doesn't interpret the CMYK data correctly (possibly stripping a color profile). This is easily verified using libjpeg standalone tools (like jpegtran that you can compile when compiling libjpeg).

The problem is probably that Photoshop does not follow the spec and
for some reason writes CMYK JPEGs with inverted inks.

Sadly there's not a (as far as I know) 100% way to spot Photoshop CMYK
JPEGs. If you think you might have a P/shop CMYK JPEG, you need to do
255-r, 255-g, 255-b, 255-k after loading it.

Here's a previous thread on the issue:

http://aspn.activestate.com/ASPN/Mail/Message/image-sig/3110476

> On the other hand, using ImageMagick python bindings (PythonMagick) worked fine (I assume ImageMagick doesn't rely on libjpeg).

IM does use libjpeg, but it guesses that if there's an Adobe app
marker it should invert inks. I don't think this is foolproof though.

John

From kevin at cazabon.com  Mon Mar  2 23:52:31 2009
From: kevin at cazabon.com (Kevin Cazabon)
Date: Mon, 2 Mar 2009 17:52:31 -0500
Subject: [Image-SIG] playing with CMYK images
In-Reply-To: <DDBEBD8A-20CB-4FEF-9EBC-A5D0174BF113@begeistert.org>
References: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
	<DDBEBD8A-20CB-4FEF-9EBC-A5D0174BF113@begeistert.org>
Message-ID: <5DBF929E-B78D-484C-A5EB-3ABC36671F5B@cazabon.com>

The jpeg plugin doesn't handle CMYK properly - I've posted patches in  
the past... trivial to fix.

Kevin.


On 02 Mar 2009, at 14:04, Charlie Clark wrote:

>
> Am 27.02.2009 um 19:09 schrieb Ask Holme:
>
>> Hey list
>>
>> i have just tried using PIL to crop and resize CMYK images (with  
>> CMYK as end result, no color mode conversion). I must say i was  
>> surprised how awfull the result look. Even though PIL is not  
>> supposed to touch the color mode, colors is totally fucked up in  
>> the modified image.
>>
>> Is that because i'm doing something wrong or is the only solution  
>> simply to switch to imagemagick if i want better CMYK support
>
> wow! As someone who normally appreciates informality on mailing  
> lists and relishes in the triumph of the vernacular I found this a  
> rude first post.
>
> Can you please provide some details as to what's gone so horribly  
> wrong. I would be surprised if the problem is actually with PIL per  
> se and not with one of the underlying C libraries.
>
> Charlie
> --
> Charlie Clark
> Helmholtzstr. 20
> D?sseldorf
> D- 40215
> Tel: +49-211-938-5360
> GSM: +49-178-782-6226
>
>
>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig


From d_berthelot at yahoo.com  Tue Mar  3 04:45:31 2009
From: d_berthelot at yahoo.com (David Berthelot)
Date: Mon, 2 Mar 2009 19:45:31 -0800 (PST)
Subject: [Image-SIG] playing with CMYK images
References: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
	<70DCC813-8501-4B29-A2EF-BD8190465A2A@timhatch.com>
	<430166.97363.qm@web52108.mail.re2.yahoo.com>
	<522c6460903021458y10929c48sba7f89cf766a6d88@mail.gmail.com>
Message-ID: <874882.7370.qm@web52104.mail.re2.yahoo.com>


Thanks for the explanation, I followed your link to the thread but I couldn't find the attachement you mentioned in it (because I'd like to use your patch !).
Is PIL still maintained (considering the patch is 3 years old) ? Or is there a specific reason it's not included in PIL ?



----- Original Message ----
From: "jcupitt at gmail.com" <jcupitt at gmail.com>
Cc: image-sig <image-sig at python.org>
Sent: Monday, March 2, 2009 2:58:04 PM
Subject: Re: [Image-SIG] playing with CMYK images

2009/3/2 David Berthelot <d_berthelot at yahoo.com>:
> I ran into a similar problems with CMYK and I think it's because libjpeg doesn't interpret the CMYK data correctly (possibly stripping a color profile). This is easily verified using libjpeg standalone tools (like jpegtran that you can compile when compiling libjpeg).

The problem is probably that Photoshop does not follow the spec and
for some reason writes CMYK JPEGs with inverted inks.

Sadly there's not a (as far as I know) 100% way to spot Photoshop CMYK
JPEGs. If you think you might have a P/shop CMYK JPEG, you need to do
255-r, 255-g, 255-b, 255-k after loading it.

Here's a previous thread on the issue:

http://aspn.activestate.com/ASPN/Mail/Message/image-sig/3110476

> On the other hand, using ImageMagick python bindings (PythonMagick) worked fine (I assume ImageMagick doesn't rely on libjpeg).

IM does use libjpeg, but it guesses that if there's an Adobe app
marker it should invert inks. I don't think this is foolproof though.

John
_______________________________________________
Image-SIG maillist  -  Image-SIG at python.org
http://mail.python.org/mailman/listinfo/image-sig



      

From tim at timhatch.com  Tue Mar  3 06:42:03 2009
From: tim at timhatch.com (Tim Hatch)
Date: Mon, 2 Mar 2009 21:42:03 -0800
Subject: [Image-SIG] playing with CMYK images
In-Reply-To: <874882.7370.qm@web52104.mail.re2.yahoo.com>
References: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
	<70DCC813-8501-4B29-A2EF-BD8190465A2A@timhatch.com>
	<430166.97363.qm@web52108.mail.re2.yahoo.com>
	<522c6460903021458y10929c48sba7f89cf766a6d88@mail.gmail.com>
	<874882.7370.qm@web52104.mail.re2.yahoo.com>
Message-ID: <33888160-BF91-49A9-AC1B-37E557F1846B@timhatch.com>

> Thanks for the explanation, I followed your link to the thread but I  
> couldn't find the attachement you mentioned in it (because I'd like  
> to use your patch !).

Try http://mail.python.org/pipermail/image-sig/2006-April/003871.html  
instead.

Tim

From ask at jillion.dk  Mon Mar  2 20:45:30 2009
From: ask at jillion.dk (Ask Holme)
Date: Mon, 2 Mar 2009 20:45:30 +0100
Subject: [Image-SIG] playing with CMYK images
In-Reply-To: <DDBEBD8A-20CB-4FEF-9EBC-A5D0174BF113@begeistert.org>
References: <fe6830130902271009h7f20230bgff7fc2b3f421df23@mail.gmail.com>
	<DDBEBD8A-20CB-4FEF-9EBC-A5D0174BF113@begeistert.org>
Message-ID: <fe6830130903021145y1f49d0a9h5461300cc3bbd1f5@mail.gmail.com>

Hey list

I'm sorry if your guys find my post rude - it wasn't meant that way!
i find PIL a great product, but when writing the post i might have been a
little bit fustrated.

I will post an example tomorrow, when i fetch it - i might note, that i
currently
have made a solution with imagemagick (running via an perl-script since i
couldn't find working python bindings to magick wand) and that works
flawlessly...

But i would like a native python solution - i'll upload examples tomorrow.


On Mon, Mar 2, 2009 at 8:04 PM, Charlie Clark <charlie at begeistert.org>wrote:

>
> Am 27.02.2009 um 19:09 schrieb Ask Holme:
>
>  Hey list
>>
>> i have just tried using PIL to crop and resize CMYK images (with CMYK as
>> end result, no color mode conversion). I must say i was surprised how awfull
>> the result look. Even though PIL is not supposed to touch the color mode,
>> colors is totally fucked up in the modified image.
>>
>> Is that because i'm doing something wrong or is the only solution simply
>> to switch to imagemagick if i want better CMYK support
>>
>
> wow! As someone who normally appreciates informality on mailing lists and
> relishes in the triumph of the vernacular I found this a rude first post.
>
> Can you please provide some details as to what's gone so horribly wrong. I
> would be surprised if the problem is actually with PIL per se and not with
> one of the underlying C libraries.
>
> Charlie
> --
> Charlie Clark
> Helmholtzstr. 20
> D?sseldorf
> D- 40215
> Tel: +49-211-938-5360
> GSM: +49-178-782-6226
>
>
>
>


-- 
Ask Holme - 28 76 98 87
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090302/5ff5c2bf/attachment.htm>

From paulegan at mail.com  Tue Mar  3 18:56:42 2009
From: paulegan at mail.com (Paul Egan)
Date: Tue, 3 Mar 2009 09:56:42 -0800 (PST)
Subject: [Image-SIG] PIL 1.1.6 and convert cmyk jpegs to rgb and writing
	png files
References: <f1e83f4f0812150053x2caec770ja2c901770978d2a@mail.gmail.com>
	<49AD6B49.4080401@yahoo.com>
Message-ID: <934271.45593.qm@web30803.mail.mud.yahoo.com>

Just in case this CMYK patch does make it into the PIL svn repo...
there's a small bug: before calling invert, self.im should be checked.
This is necessary because the ImageFile.Parser.feed method tries an 
open() to see if the image is complete.

Attached is an updated patch.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PIL.JpegImagePlugin.diff
Type: text/x-diff
Size: 2154 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090303/08705f12/attachment.diff>

From chris at stemcel.co.uk  Wed Mar  4 14:06:12 2009
From: chris at stemcel.co.uk (Chris Adams)
Date: Wed, 4 Mar 2009 13:06:12 +0000
Subject: [Image-SIG] Using PIL to merge postscript generated text and png
	files
Message-ID: <67A4B838-D111-47B0-AA83-DA217BADFE00@stemcel.co.uk>

Hi there,

I'm trying to programatically generate text onto an existing png file  
as part of a very simple online tool for an NGO, and I'm looking at  
using PIL to do this.

Blank badge:
http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/

The copy  would be pasted in goes here:
http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/posted_copy_badge.png

And the rollover state would be generated as well here:
http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/posted_copy_badge_rollover.png

However, the issue here is that I'm trying to do this with a  
Postscript font (not a Truetype .ttf one), which needs the PSDraw  
class, rather than the basic ImageFont class, and I was under the  
impression that it was designed for printing rather than output for  
the screen.

I'm used to using GUI apps for design related work like this or  
writing html and css on the front end of websites, and I'm very new to  
python, but I naively assumed that doing this would be simple, and  
it's proving not to be the case so far.

What's the best way to try doing what I'm doing?






-- 
---
I'm currently only checking my email at 9am, midday and at 4pm.
If you need a response from me urgently, please call or text my  
mobile, or contact me via Skype (chris.d.adams).
---
Chris Adams
Stemcel Studios
The Hub
5 Torrens Street
London
EC1V 1NQ

email: chris at stemcel.co.uk
web:  www.stemcel.co.uk
twitter:chris_d_adams
skype: chris.d.adams
mob: 07974 368 229
tel: 0207 558 8971


From cannon.el at gmail.com  Wed Mar  4 17:26:55 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Wed, 4 Mar 2009 08:26:55 -0800
Subject: [Image-SIG] Using PIL to merge postscript generated text and
	png files
In-Reply-To: <67A4B838-D111-47B0-AA83-DA217BADFE00@stemcel.co.uk>
References: <67A4B838-D111-47B0-AA83-DA217BADFE00@stemcel.co.uk>
Message-ID: <b986b290903040826n37e9f15dpebcbe2a40df0c446@mail.gmail.com>

My best suggestion is to convert the font you need to a format that
PIL can understand, either its own bitmap format or to a truetype
font. Font conversion utilities exist, and you would only need to do
the conversion once, and then just use the converted font. It may not
be completely elegant, but it is probably the minimum amount of work.
Edward

On Wed, Mar 4, 2009 at 5:06 AM, Chris Adams <chris at stemcel.co.uk> wrote:
> Hi there,
>
> I'm trying to programatically generate text onto an existing png file as
> part of a very simple online tool for an NGO, and I'm looking at using PIL
> to do this.
>
> Blank badge:
> http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/
>
> The copy ?would be pasted in goes here:
> http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/posted_copy_badge.png
>
> And the rollover state would be generated as well here:
> http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/posted_copy_badge_rollover.png
>
> However, the issue here is that I'm trying to do this with a Postscript font
> (not a Truetype .ttf one), which needs the PSDraw class, rather than the
> basic ImageFont class, and I was under the impression that it was designed
> for printing rather than output for the screen.
>
> I'm used to using GUI apps for design related work like this or writing html
> and css on the front end of websites, and I'm very new to python, but I
> naively assumed that doing this would be simple, and it's proving not to be
> the case so far.
>
> What's the best way to try doing what I'm doing?
>
>
>
>
>
>
> --
> ---
> I'm currently only checking my email at 9am, midday and at 4pm.
> If you need a response from me urgently, please call or text my mobile, or
> contact me via Skype (chris.d.adams).
> ---
> Chris Adams
> Stemcel Studios
> The Hub
> 5 Torrens Street
> London
> EC1V 1NQ
>
> email: chris at stemcel.co.uk
> web: ?www.stemcel.co.uk
> twitter:chris_d_adams
> skype: chris.d.adams
> mob: 07974 368 229
> tel: 0207 558 8971
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From wichert at simplon.biz  Tue Mar  3 22:21:05 2009
From: wichert at simplon.biz (Wichert Akkerman)
Date: Tue, 03 Mar 2009 22:21:05 +0100
Subject: [Image-SIG] OverflowError on image save
Message-ID: <49AD9F41.2050604@simplon.biz>

I have a small routine which tries to verify if an uploaded file is a 
valid image. It works like this:

         image_data=StringIO(data)
         try:
             pil_data=PIL.Image.open(image_data)
         except IOError: # PIL abuses IOError to report parsing errors
             raise Invalid(self.message("bad_image", state),
                     value, state)

         pil_data.load()
         (width,height)=pil_data.size


For some images I get an OverflowError on the load() call:

ModulePIL.ImageFile:*155*  in|load|
>>     <imap://wichert%40simplon%2Ebiz at mx1.simplon.biz:993/fetch%3EUID%3E.2style4you.Site%20Errors%3E331#>self.load_prepare()
ModulePIL.ImageFile:*223*  in|load_prepare|
>>     <imap://wichert%40simplon%2Ebiz at mx1.simplon.biz:993/fetch%3EUID%3E.2style4you.Site%20Errors%3E331#>self.im  =  Image.core.new(self.mode,  self.size)*
OverflowError: long int too large to convert to int
*

I can not find any information as to what the problem might be. It seems 
to only occur for bmp files; jpeg and png appear to work correctly.

This is using PIL 1.1.6 and Python 2.5 on a Linux system.

Wichert.
<< 
<imap://wichert%40simplon%2Ebiz at mx1.simplon.biz:993/fetch%3EUID%3E.2style4you.Site%20Errors%3E331#> 
self.map = None

self.load_prepare()

# look for read/seek overrides
<< 
<imap://wichert%40simplon%2Ebiz at mx1.simplon.biz:993/fetch%3EUID%3E.2style4you.Site%20Errors%3E331#> 
if not self.im or\
self.im.mode != self.mode or self.im.size != self.size:
self.im = Image.core.new(self.mode, self.size)
# create palette (optional)
if self.mode == "P":
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090303/b66f827b/attachment.htm>

From jsh2134 at gmail.com  Wed Mar  4 01:52:39 2009
From: jsh2134 at gmail.com (Jeff Hull)
Date: Tue, 3 Mar 2009 19:52:39 -0500
Subject: [Image-SIG] Image to Matrix to Image.
Message-ID: <807450d00903031652s30112838sc5eb3b3ecbac6866@mail.gmail.com>

I am having the same problem. I can not save this image after i edit its
data.

im = myimage.load()

...do stuff to im...

im.save('jeff', "JPEG")


Traceback (most recent call last):
  File "numberfinder.py", line 116, in <module>
    main()
  File "numberfinder.py", line 97, in main
    imL = applyMask(imL, immask)
  File "numberfinder.py", line 113, in applyMask
    return im.save("jeff", "JPEG")
AttributeError: 'PixelAccess' object has no attribute 'save'


what do i convert the PixelAccess object to so that I can save it?


thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090303/7c4751e1/attachment.htm>

From cannon.el at gmail.com  Thu Mar  5 17:00:05 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Thu, 5 Mar 2009 08:00:05 -0800
Subject: [Image-SIG] Using PIL to merge postscript generated text and
	png files
In-Reply-To: <7387B742-F906-4D44-9AF7-36B50B8ACD14@stemcel.co.uk>
References: <67A4B838-D111-47B0-AA83-DA217BADFE00@stemcel.co.uk>
	<b986b290903040826n37e9f15dpebcbe2a40df0c446@mail.gmail.com>
	<7387B742-F906-4D44-9AF7-36B50B8ACD14@stemcel.co.uk>
Message-ID: <b986b290903050800y60300202xf186954416aa2787@mail.gmail.com>

As far as I know, line wrapping in PIL has to be done automatically.
You need to use the font.getsize(text) method in ImageFont it returns
a (width, height) tuple for the given text. A little work will allow
you to write a function to split the lines at the appropriate places
(I would split the text into words with the .split() string method)
and then you can draw each line individually.
Edward

On 3/5/09, Chris Adams <chris at stemcel.co.uk> wrote:
> Hi Edward, thanks for the heads up!
>
>  In the end, I used FontForge to create a Truetype font from the postscript,
> and it worked pretty seamlessly - it's not quite photoshop, but it IS free
> and cross platform (http://fontforge.sourceforge.net/)
>
>  Initially, the typefaces looked nasty and pixellated when I was using an
> initial png8 graphic -
>   instead of a full colour version.
>
>  I resolved this  case by exporting the file from photoshop as a 24bit png
> instead, which took the rough edges off the fonts.
>
>  However, I'm still stuck with the issue of having no line wrapping.
>
>  Is this something PIL can handle as well?
>
>  C
>
>
>  On 4 Mar 2009, at 16:26, Laura & Edward Cannon wrote:
>
>
> > My best suggestion is to convert the font you need to a format that
> > PIL can understand, either its own bitmap format or to a truetype
> > font. Font conversion utilities exist, and you would only need to do
> > the conversion once, and then just use the converted font. It may not
> > be completely elegant, but it is probably the minimum amount of work.
> > Edward
> >
> > On Wed, Mar 4, 2009 at 5:06 AM, Chris Adams <chris at stemcel.co.uk> wrote:
> >
> > > Hi there,
> > >
> > > I'm trying to programatically generate text onto an existing png file as
> > > part of a very simple online tool for an NGO, and I'm looking at using
> PIL
> > > to do this.
> > >
> > > Blank badge:
> > >
> http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/
> > >
> > > The copy  would be pasted in goes here:
> > >
> http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/posted_copy_badge.png
> > >
> > > And the rollover state would be generated as well here:
> > >
> http://test.stemcel.co.uk/clients/org/org-widget/blank_badge/posted_copy_badge_rollover.png
> > >
> > > However, the issue here is that I'm trying to do this with a Postscript
> font
> > > (not a Truetype .ttf one), which needs the PSDraw class, rather than the
> > > basic ImageFont class, and I was under the impression that it was
> designed
> > > for printing rather than output for the screen.
> > >
> > > I'm used to using GUI apps for design related work like this or writing
> html
> > > and css on the front end of websites, and I'm very new to python, but I
> > > naively assumed that doing this would be simple, and it's proving not to
> be
> > > the case so far.
> > >
> > > What's the best way to try doing what I'm doing?
> > >
> > >
> > >
> > >
> > >
> > >
> > > --
> > > ---
> > > I'm currently only checking my email at 9am, midday and at 4pm.
> > > If you need a response from me urgently, please call or text my mobile,
> or
> > > contact me via Skype (chris.d.adams).
> > > ---
> > > Chris Adams
> > > Stemcel Studios
> > > The Hub
> > > 5 Torrens Street
> > > London
> > > EC1V 1NQ
> > >
> > > email: chris at stemcel.co.uk
> > > web:  www.stemcel.co.uk
> > > twitter:chris_d_adams
> > > skype: chris.d.adams
> > > mob: 07974 368 229
> > > tel: 0207 558 8971
> > >
> > > _______________________________________________
> > > Image-SIG maillist  -  Image-SIG at python.org
> > > http://mail.python.org/mailman/listinfo/image-sig
> > >
> > >
> > _______________________________________________
> > Image-SIG maillist  -  Image-SIG at python.org
> > http://mail.python.org/mailman/listinfo/image-sig
> >
>
>  --
>  ---
>  I'm currently only checking my email at 9am, midday and at 4pm.
>  If you need a response from me urgently, please call or text my mobile, or
> contact me via Skype (chris.d.adams).
>  ---
>  Chris Adams
>  Stemcel Studios
>  The Hub
>  5 Torrens Street
>  London
>  EC1V 1NQ
>
>  email: chris at stemcel.co.uk
>  web:  www.stemcel.co.uk
>  twitter:chris_d_adams
>  skype: chris.d.adams
>  mob: 07974 368 229
>  tel: 0207 558 8971
>
>

From fredrik at pythonware.com  Thu Mar  5 18:41:20 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 5 Mar 2009 18:41:20 +0100
Subject: [Image-SIG] OverflowError on image save
In-Reply-To: <49AD9F41.2050604@simplon.biz>
References: <49AD9F41.2050604@simplon.biz>
Message-ID: <368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>

Can you print the following attributes for a broken file, before calling load:

    pil_data.mode
    pil_data.size
    pil_data.tile

</F>

2009/3/3 Wichert Akkerman <wichert at simplon.biz>:
> I have a small routine which tries to verify if an uploaded file is a valid
> image. It works like this:
>
>         image_data=StringIO(data)
>         try:
>             pil_data=PIL.Image.open(image_data)
>         except IOError: # PIL abuses IOError to report parsing errors
>             raise Invalid(self.message("bad_image", state),
>                     value, state)
>
>         pil_data.load()
>         (width,height)=pil_data.size
>
> For some images I get an OverflowError on the load() call:
>
> Module PIL.ImageFile:155 in load
>>>? self.load_prepare()
> Module PIL.ImageFile:223 in load_prepare
>>>? self.im = Image.core.new(self.mode, self.size)
> OverflowError: long int too large to convert to int
>
> I can not find any information as to what the problem might be. It seems to
> only occur for bmp files; jpeg and png appear to work correctly.
>
> This is using PIL 1.1.6 and Python 2.5 on a Linux system.
>
> Wichert.
> <<? ??????????????????? self.map = None
>
> ?????????????? self.load_prepare()
>
> ?????????????? # look for read/seek overrides
> <<? ??????? if not self.im or\
> ????????????????? self.im.mode != self.mode or self.im.size != self.size:
> ?????????????????? self.im = Image.core.new(self.mode, self.size)
> ?????????????? # create palette (optional)
> ?????????????? if self.mode == "P":
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>

From fredrik at pythonware.com  Thu Mar  5 19:27:44 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 5 Mar 2009 19:27:44 +0100
Subject: [Image-SIG] Image to Matrix to Image.
In-Reply-To: <807450d00903031652s30112838sc5eb3b3ecbac6866@mail.gmail.com>
References: <807450d00903031652s30112838sc5eb3b3ecbac6866@mail.gmail.com>
Message-ID: <368a5cd50903051027v2357842ft449d20ae987c505b@mail.gmail.com>

2009/3/4 Jeff Hull <jsh2134 at gmail.com>:

> I am having the same problem. I can not save this image after i edit its
> data.
>
> im = myimage.load()
>
> ...do stuff to im...
>
> im.save('jeff', "JPEG")

make that

   myimage.save(...)

the object that's returned from load is a reference to the contents of
the image, not the image itself.

</F>

From jorge-m-pereira at ptinovacao.pt  Thu Mar  5 15:40:58 2009
From: jorge-m-pereira at ptinovacao.pt (Jorge Pereira)
Date: Thu, 5 Mar 2009 14:40:58 +0000
Subject: [Image-SIG] ZLIB (PNG/ZIP) support not available
Message-ID: <FE488BAD8B4EBB4BBEC64B3EEA4497E6607C2FE9D8@INOAVREX11.ptin.corpPT.com>

Hi,
We're trying to install PIL 1.1.6 on a RHEL 4Up4 x64, but we are having some difficulties.
Although zlib RHEL's rpms are installed (zlib-1.2.1.2-1.2.i386 and zlib-1.2.1.2-1.2.x86_64), when executing "python setup.py build_ext -I", we keep getting the message saying that ZLIB (PNG/ZIP) support is not available. Same happens with FREETYPE and JPEG.
Can you hel?

Best regards


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090305/d64e9624/attachment.htm>

From Tim.Swan at disney.com  Thu Mar  5 19:23:57 2009
From: Tim.Swan at disney.com (Swan, Tim)
Date: Thu, 5 Mar 2009 18:23:57 +0000
Subject: [Image-SIG] PIL 1.1.6, python 64bit on Windows XP x64
Message-ID: <918738C226E29C488423AC1B6D203D2201B248FF@SM-GBLO-VXM31A.emea.wdpr.disney.com>


Hi,

I've been trying to get PIL 1.1.6 working on python 64bit:
Python version : 2.5.2 (r252:60911, Feb 21 2008, 13:17:27) [MSC v.1400
64 bit (AMD64)]
OS : Windows XP Professional x64 SP2.

The installer on the main download page appears to only be valid for
32bit python, causing a variety of import errors when used with 64bit
python.
Is there any plan to provide an installer for 64bit versions of python
on 64 bit Windows in the future?

I've sucessfully built a version locally with DevStudio 2005, including
zlib and jpeg-6b support. I'll detail the steps required later for
posterity, it may be useful for someone browsing this forum at a later
time - I have no plans to provide an installer myself!

 I did have to work around an issue, which I thought was worth raising
to you folks, in case there was a better way to fix this.
After building locally (setup.py build, setup.py install) and importing
PIL.Image, the quick test detailed below caused my python process to
crash.

Import PIL.Image
Im = PIL.Image.open('my_image.gif')
Im2 = im.rotate(45)
Im2.save()

At the point of saving, the process crashes with a windows IO exception.
Digging into the python, and then into the C extension code, I found the
problem appears to be due to some conflict with using the file
descriptor that's being passed into _encode_to_file (encode.c) from
_save in ImageFile.py (line 492 in ImageFile.py in the source I'm
looking at). After trying various different things, I found that I could
not use the file descriptor int (acquired via fp.fileno(), line 469 in
ImageFile.py) passed into encode_to_file - trying to write to a file
using that file descriptor was causing the crash (for example, line 164
in encode.c). I tried using _rdinfo to convert the file descriptor to a
FILE pointer, and work with that, but that crashed in _rdinfo, and I
tried a couple of different conversions using calls like _get_osfhandle
, but pretty much everything I tried continued to cause crashes. My
knowledge of Windows is pretty minimal, so I've no idea why using that
file descriptor should be problematic (and, presumably, works just fine
with the 32bit python version).

After taking a step back, I realised that the code in ImageFile has two
paths, and by putting a forced call to raise an AttributeError around
line 469 caused the _save function to take the top code path, and this
works just fine for me. I ended up modifying the code to look like this:

Line 467, ImageFile.py
    try:
        # FIXME: force an AttributeError exception to use the python
file-compatible object path
        raise AttributeError
        fh = fp.fileno()
        fp.flush()
    except AttributeError:
	...

It's not clean, but it appears to work. Anyone got any ideas why? Can
someone suggest a cleaner way of doing this?




If anyone is interested, here's the steps I followed to get PIL 1.1.6
working locally for me - DevStudio 2005 required.

Step 1: unarchive the source to a location locally, I'll refer to this
as the build root directory (in my case, c:\temp\Imaging-1.1.6)

Step 2: run the visual studio 2005 command prompt for the relevant bits
of the OS
(in my case from Windows start menu: Programs -> (somewhere) ->
Microsoft Visual Studio 2005 -> Visual Studio Tools -> Visual Studio
2005 x64 Win64 Command Prompt
http://docs.python.org/dist/module-distutils.msvccompiler.html

Step 3: set up the following environment variables at the command prompt
set DISTUTILS_USE_SDK=1
set MSSdk=1

Step 4: edit ImageFile.py as described above

Step 5: (optional) edit setup.py to point to the correct locations for
zlib, jpeg-6b, etc. source and built libs

Step 6 : build PIL using "setup.py build" in the build root directory.
(it's worth at this point confirming that you're running setup.py with
the correct (64bit) version of python - slap a pdb break in there and
check sys.version if needed)

Step 7 : embed the manifests for the two DLLS by running mt.exe in the
build root directory:
Mt.exe -manifest build\lib.win32-2.5_imaging.pyd.manifest
-outputresource:build\lib.win32-2.5_imaging.pyd;2
Mt.exe -manifest build\lib.win32-2.5_imagingmath.pyd.manifest
-outputresource:build\lib.win32-2.5_imagingmath.pyd;2
http://msdn.microsoft.com/en-us/library/ms235591(VS.80).aspx

Step 8: install PIL (setup.py install)


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090305/f64e0c06/attachment.htm>

From karsten.hiddemann at mathematik.uni-dortmund.de  Thu Mar  5 21:49:54 2009
From: karsten.hiddemann at mathematik.uni-dortmund.de (Karsten Hiddemann)
Date: Thu, 05 Mar 2009 21:49:54 +0100
Subject: [Image-SIG] ZLIB (PNG/ZIP) support not available
In-Reply-To: <FE488BAD8B4EBB4BBEC64B3EEA4497E6607C2FE9D8@INOAVREX11.ptin.corpPT.com>
References: <FE488BAD8B4EBB4BBEC64B3EEA4497E6607C2FE9D8@INOAVREX11.ptin.corpPT.com>
Message-ID: <49B03AF2.5040700@mathematik.uni-dortmund.de>

Jorge Pereira schrieb:
> We?re trying to install PIL 1.1.6 on a RHEL 4Up4 x64, but we are having
> some difficulties.
> 
> Although zlib RHEL?s rpms are installed (zlib-1.2.1.2-1.2.i386 and
> zlib-1.2.1.2-1.2.x86_64), when executing ?python setup.py build_ext ?I?,
> we keep getting the message saying that ZLIB (PNG/ZIP) support is not
> available. Same happens with FREETYPE and JPEG.

Same answer as for the other people with these kinds of problems:

Install the development packages of the libs which include the sources.

Cheers,
Karsten

From fredrik at pythonware.com  Thu Mar  5 22:04:38 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 5 Mar 2009 22:04:38 +0100
Subject: [Image-SIG] ZLIB (PNG/ZIP) support not available
In-Reply-To: <FE488BAD8B4EBB4BBEC64B3EEA4497E6607C2FE9D8@INOAVREX11.ptin.corpPT.com>
References: <FE488BAD8B4EBB4BBEC64B3EEA4497E6607C2FE9D8@INOAVREX11.ptin.corpPT.com>
Message-ID: <368a5cd50903051304n5d00195fn29d27cadc3d5b8c6@mail.gmail.com>

I'm not up to date with RHEL packaging, but judging from e.g.

    http://rpm.pbone.net/index.php3/stat/4/idpl/5602905/com/zlib-1.2.1.2-1.2.x86_64.rpm.html

the packages you have are runtime libraries, not development packages
(i.e. header files and link libraries are missing).

Look for packages with -dev or -devel in their names.  Or look for a
python-imaging package from RedHat; I'm pretty sure you can find
prebuilt packages for all major Linuxes these days.

Cheers /F

2009/3/5 Jorge Pereira <jorge-m-pereira at ptinovacao.pt>:
> Hi,
>
> We?re trying to install PIL 1.1.6 on a RHEL 4Up4 x64, but we are having some
> difficulties.
>
> Although zlib RHEL?s rpms are installed (zlib-1.2.1.2-1.2.i386 and
> zlib-1.2.1.2-1.2.x86_64), when executing ?python setup.py build_ext ?I?, we
> keep getting the message saying that ZLIB (PNG/ZIP) support is not
> available. Same happens with FREETYPE and JPEG.
>
> Can you hel?
>
>
>
> Best regards
>
>
>
>
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>

From tim at timhatch.com  Thu Mar  5 22:03:24 2009
From: tim at timhatch.com (Tim Hatch)
Date: Thu, 5 Mar 2009 13:03:24 -0800
Subject: [Image-SIG] ZLIB (PNG/ZIP) support not available
In-Reply-To: <FE488BAD8B4EBB4BBEC64B3EEA4497E6607C2FE9D8@INOAVREX11.ptin.corpPT.com>
References: <FE488BAD8B4EBB4BBEC64B3EEA4497E6607C2FE9D8@INOAVREX11.ptin.corpPT.com>
Message-ID: <699AF96D-0185-49FD-8310-49534C956688@timhatch.com>

> Although zlib RHEL?s rpms are installed (zlib-1.2.1.2-1.2.i386 and  
> zlib-1.2.1.2-1.2.x86_64)...

Make sure you have the dev versions of those packages as well, namely  
zlib-devel, freetype-devel, etc.

Tim

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090305/360f300d/attachment.htm>

From fredrik at pythonware.com  Thu Mar  5 22:09:42 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 5 Mar 2009 22:09:42 +0100
Subject: [Image-SIG] PIL 1.1.6, python 64bit on Windows XP x64
In-Reply-To: <918738C226E29C488423AC1B6D203D2201B248FF@SM-GBLO-VXM31A.emea.wdpr.disney.com>
References: <918738C226E29C488423AC1B6D203D2201B248FF@SM-GBLO-VXM31A.emea.wdpr.disney.com>
Message-ID: <368a5cd50903051309t4530ac7cr772a9d40f7cfbd53@mail.gmail.com>

Thanks for the howto.  Now I just need to go find myself a 64-bit box
to play with ;-)

As for the file descriptor crashes, things like that is usually caused
by a mismatch between the C runtime libraries used for core Python and
the extension (the file handle is allocated by the Python main
process, but used by code in the extension.  If the internal structs
don't match, all sort of oddities can happen).  You can use e.g.
dumpbin /imports on the python DLL and the PIL PYD's, and verify that
they're not importing differenct MSVC-something DLL:s.

</F>

2009/3/5 Swan, Tim <Tim.Swan at disney.com>:
>
> Hi,
>
> I've been trying to get PIL 1.1.6 working on python 64bit:
> Python version : 2.5.2 (r252:60911, Feb 21 2008, 13:17:27) [MSC v.1400 64
> bit (AMD64)]
> OS : Windows XP Professional x64 SP2.
>
> The installer on the main download page appears to only be valid for 32bit
> python, causing a variety of import errors when used with 64bit python.
>
> Is there any plan to provide an installer for 64bit versions of python on 64
> bit Windows in the future?
>
> I've sucessfully built a version locally with DevStudio 2005, including zlib
> and jpeg-6b support. I'll detail the steps required later for posterity, it
> may be useful for someone browsing this forum at a later time - I have no
> plans to provide an installer myself!
>
> ?I did have to work around an issue, which I thought was worth raising to
> you folks, in case there was a better way to fix this.
>
> After building locally (setup.py build, setup.py install) and importing
> PIL.Image, the quick test detailed below caused my python process to crash.
>
> Import PIL.Image
> Im = PIL.Image.open('my_image.gif')
> Im2 = im.rotate(45)
> Im2.save()
>
> At the point of saving, the process crashes with a windows IO exception.
> Digging into the python, and then into the C extension code, I found the
> problem appears to be due to some conflict with using the file descriptor
> that's being passed into _encode_to_file (encode.c) from _save in
> ImageFile.py (line 492 in ImageFile.py in the source I'm looking at). After
> trying various different things, I found that I could not use the file
> descriptor int (acquired via fp.fileno(), line 469 in ImageFile.py) passed
> into encode_to_file - trying to write to a file using that file descriptor
> was causing the crash (for example, line 164 in encode.c). I tried using
> _rdinfo to convert the file descriptor to a FILE pointer, and work with
> that, but that crashed in _rdinfo, and I tried a couple of different
> conversions using calls like _get_osfhandle
>
> , but pretty much everything I tried continued to cause crashes. My
> knowledge of Windows is pretty minimal, so I've no idea why using that file
> descriptor should be problematic (and, presumably, works just fine with the
> 32bit python version).
>
> After taking a step back, I realised that the code in ImageFile has two
> paths, and by putting a forced call to raise an AttributeError around line
> 469 caused the _save function to take the top code path, and this works just
> fine for me. I ended up modifying the code to look like this:
>
> Line 467, ImageFile.py
> ??? try:
> ??????? # FIXME: force an AttributeError exception to use the python
> file-compatible object path
> ??????? raise AttributeError
> ??????? fh = fp.fileno()
> ??????? fp.flush()
> ??? except AttributeError:
> ??????? ...
>
> It's not clean, but it appears to work. Anyone got any ideas why? Can
> someone suggest a cleaner way of doing this?
>
>
> If anyone is interested, here's the steps I followed to get PIL 1.1.6
> working locally for me - DevStudio 2005 required.
>
> Step 1: unarchive the source to a location locally, I'll refer to this as
> the build root directory (in my case, c:\temp\Imaging-1.1.6)
>
> Step 2: run the visual studio 2005 command prompt for the relevant bits of
> the OS
> (in my case from Windows start menu: Programs -> (somewhere) -> Microsoft
> Visual Studio 2005 -> Visual Studio Tools -> Visual Studio 2005 x64 Win64
> Command Prompt
>
> http://docs.python.org/dist/module-distutils.msvccompiler.html
>
> Step 3: set up the following environment variables at the command prompt
> set DISTUTILS_USE_SDK=1
> set MSSdk=1
>
> Step 4: edit ImageFile.py as described above
>
> Step 5: (optional) edit setup.py to point to the correct locations for zlib,
> jpeg-6b, etc. source and built libs
>
> Step 6 : build PIL using "setup.py build" in the build root directory.
> (it's worth at this point confirming that you're running setup.py with the
> correct (64bit) version of python - slap a pdb break in there and check
> sys.version if needed)
>
> Step 7 : embed the manifests for the two DLLS by running mt.exe in the build
> root directory:
> Mt.exe -manifest build\lib.win32-2.5_imaging.pyd.manifest
> -outputresource:build\lib.win32-2.5_imaging.pyd;2
> Mt.exe -manifest build\lib.win32-2.5_imagingmath.pyd.manifest
> -outputresource:build\lib.win32-2.5_imagingmath.pyd;2
> http://msdn.microsoft.com/en-us/library/ms235591(VS.80).aspx
>
> Step 8: install PIL (setup.py install)
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>

From wichert at simplon.biz  Fri Mar  6 14:50:11 2009
From: wichert at simplon.biz (Wichert Akkerman)
Date: Fri, 06 Mar 2009 14:50:11 +0100
Subject: [Image-SIG] OverflowError on image save
In-Reply-To: <368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>
References: <49AD9F41.2050604@simplon.biz>
	<368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>
Message-ID: <49B12A13.2080406@simplon.biz>

The first image:

 >>> i=PIL.Image.open(open("/Users/wichert/Desktop/dont.bmp"))
 >>> i
<PIL.BmpImagePlugin.BmpImageFile instance at 0x43f490>
 >>> i.mode
'RGB'
 >>> i.size
(290, 4294967075L)
 >>> i.tile
[('raw', (0, 0, 290, 4294967075L), 54, ('BGR', 872, -1))]

and the second image:

 >>> i=PIL.Image.open(open("/Users/wichert/Desktop/dont2.bmp"))
 >>> i.mode
'RGB'
 >>> i.size
(288, 4294967078L)
 >>> i.tile
[('raw', (0, 0, 288, 4294967078L), 54, ('BGR', 864, -1))]


Regards,
Wichert.

On 3/5/09 6:41 PM, Fredrik Lundh wrote:
> Can you print the following attributes for a broken file, before calling load:
>
>      pil_data.mode
>      pil_data.size
>      pil_data.tile
>
> </F>
>
> 2009/3/3 Wichert Akkerman<wichert at simplon.biz>:
>    
>> I have a small routine which tries to verify if an uploaded file is a valid
>> image. It works like this:
>>
>>          image_data=StringIO(data)
>>          try:
>>              pil_data=PIL.Image.open(image_data)
>>          except IOError: # PIL abuses IOError to report parsing errors
>>              raise Invalid(self.message("bad_image", state),
>>                      value, state)
>>
>>          pil_data.load()
>>          (width,height)=pil_data.size
>>
>> For some images I get an OverflowError on the load() call:
>>
>> Module PIL.ImageFile:155 in load
>>      
>>>>    self.load_prepare()
>>>>          
>> Module PIL.ImageFile:223 in load_prepare
>>      
>>>>    self.im = Image.core.new(self.mode, self.size)
>>>>          
>> OverflowError: long int too large to convert to int
>>
>> I can not find any information as to what the problem might be. It seems to
>> only occur for bmp files; jpeg and png appear to work correctly.
>>
>> This is using PIL 1.1.6 and Python 2.5 on a Linux system.
>>
>> Wichert.
>> <<                       self.map = None
>>
>>                 self.load_prepare()
>>
>>                 # look for read/seek overrides
>> <<           if not self.im or\
>>                    self.im.mode != self.mode or self.im.size != self.size:
>>                     self.im = Image.core.new(self.mode, self.size)
>>                 # create palette (optional)
>>                 if self.mode == "P":
>> _______________________________________________
>> Image-SIG maillist  -  Image-SIG at python.org
>> http://mail.python.org/mailman/listinfo/image-sig
>>
>>
>>      


From fredrik at pythonware.com  Fri Mar  6 17:43:46 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 6 Mar 2009 17:43:46 +0100
Subject: [Image-SIG] OverflowError on image save
In-Reply-To: <29bf24a80903060752h31c2fdd4y6f3114539a178ad@mail.gmail.com>
References: <49AD9F41.2050604@simplon.biz>
	<368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>
	<49B12A13.2080406@simplon.biz>
	<29bf24a80903060752h31c2fdd4y6f3114539a178ad@mail.gmail.com>
Message-ID: <368a5cd50903060843y1dfbdb07u89a935941b5e417@mail.gmail.com>

Does anyone have a (preferrably) small example, or a simple way to
create a broken file on a stock Windows XP machine?  An URL or an
off-list mail works fine.

</F>

On Fri, Mar 6, 2009 at 4:52 PM, Howard Lightstone
<howard at eegsoftware.com> wrote:
> This LOOKS like a bug in Windows XP which creates .BMP files via DirectX
> with an incorrect bitmap header size (54 vice 40).? Changing the 54 to 40
> makes the file readable again.? Most utilities seem to ignore this but
> Windows itself won't be able to 'read' it as a .BMP file.
>
> (using XP SP2 and DirectX 9.0c ... I haven't tried this on Vista yet)
>
> On Fri, Mar 6, 2009 at 5:50 AM, Wichert Akkerman <wichert at simplon.biz>
> wrote:
>>
>> The first image:
>>
>> >>> i=PIL.Image.open(open("/Users/wichert/Desktop/dont.bmp"))
>> >>> i
>> <PIL.BmpImagePlugin.BmpImageFile instance at 0x43f490>
>> >>> i.mode
>> 'RGB'
>> >>> i.size
>> (290, 4294967075L)
>> >>> i.tile
>> [('raw', (0, 0, 290, 4294967075L), 54, ('BGR', 872, -1))]
>>
>> and the second image:
>>
>> >>> i=PIL.Image.open(open("/Users/wichert/Desktop/dont2.bmp"))
>> >>> i.mode
>> 'RGB'
>> >>> i.size
>> (288, 4294967078L)
>> >>> i.tile
>> [('raw', (0, 0, 288, 4294967078L), 54, ('BGR', 864, -1))]
>>
>>
>> Regards,
>> Wichert.
>>
>> On 3/5/09 6:41 PM, Fredrik Lundh wrote:
>>>
>>> Can you print the following attributes for a broken file, before calling
>>> load:
>>>
>>> ? ? pil_data.mode
>>> ? ? pil_data.size
>>> ? ? pil_data.tile
>>>
>>> </F>
>>>
>>> 2009/3/3 Wichert Akkerman<wichert at simplon.biz>:
>>>
>>>>
>>>> I have a small routine which tries to verify if an uploaded file is a
>>>> valid
>>>> image. It works like this:
>>>>
>>>> ? ? ? ? image_data=StringIO(data)
>>>> ? ? ? ? try:
>>>> ? ? ? ? ? ? pil_data=PIL.Image.open(image_data)
>>>> ? ? ? ? except IOError: # PIL abuses IOError to report parsing errors
>>>> ? ? ? ? ? ? raise Invalid(self.message("bad_image", state),
>>>> ? ? ? ? ? ? ? ? ? ? value, state)
>>>>
>>>> ? ? ? ? pil_data.load()
>>>> ? ? ? ? (width,height)=pil_data.size
>>>>
>>>> For some images I get an OverflowError on the load() call:
>>>>
>>>> Module PIL.ImageFile:155 in load
>>>>
>>>>>>
>>>>>> ? self.load_prepare()
>>>>>>
>>>>
>>>> Module PIL.ImageFile:223 in load_prepare
>>>>
>>>>>>
>>>>>> ? self.im = Image.core.new(self.mode, self.size)
>>>>>>
>>>>
>>>> OverflowError: long int too large to convert to int
>>>>
>>>> I can not find any information as to what the problem might be. It seems
>>>> to
>>>> only occur for bmp files; jpeg and png appear to work correctly.
>>>>
>>>> This is using PIL 1.1.6 and Python 2.5 on a Linux system.
>>>>
>>>> Wichert.
>>>> << ? ? ? ? ? ? ? ? ? ? ? self.map = None
>>>>
>>>> ? ? ? ? ? ? ? ?self.load_prepare()
>>>>
>>>> ? ? ? ? ? ? ? ?# look for read/seek overrides
>>>> << ? ? ? ? ? if not self.im or\
>>>> ? ? ? ? ? ? ? ? ? self.im.mode != self.mode or self.im.size !=
>>>> self.size:
>>>> ? ? ? ? ? ? ? ? ? ?self.im = Image.core.new(self.mode, self.size)
>>>> ? ? ? ? ? ? ? ?# create palette (optional)
>>>> ? ? ? ? ? ? ? ?if self.mode == "P":
>>>> _______________________________________________
>>>> Image-SIG maillist ?- ?Image-SIG at python.org
>>>> http://mail.python.org/mailman/listinfo/image-sig
>>>>
>>>>
>>>>
>>
>> _______________________________________________
>> Image-SIG maillist ?- ?Image-SIG at python.org
>> http://mail.python.org/mailman/listinfo/image-sig
>>
>
>
>
> --
> Howard Lightstone
> hlightstone at gmail.com
> howard at eegsoftware.com
>

From howard at eegsoftware.com  Fri Mar  6 16:52:57 2009
From: howard at eegsoftware.com (Howard Lightstone)
Date: Fri, 6 Mar 2009 07:52:57 -0800
Subject: [Image-SIG] OverflowError on image save
In-Reply-To: <49B12A13.2080406@simplon.biz>
References: <49AD9F41.2050604@simplon.biz>
	<368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>
	<49B12A13.2080406@simplon.biz>
Message-ID: <29bf24a80903060752h31c2fdd4y6f3114539a178ad@mail.gmail.com>

This LOOKS like a bug in Windows XP which creates .BMP files via DirectX
with an incorrect bitmap header size (54 vice 40).  Changing the 54 to 40
makes the file readable again.  Most utilities seem to ignore this but
Windows itself won't be able to 'read' it as a .BMP file.

(using XP SP2 and DirectX 9.0c ... I haven't tried this on Vista yet)

On Fri, Mar 6, 2009 at 5:50 AM, Wichert Akkerman <wichert at simplon.biz>wrote:

> The first image:
>
> >>> i=PIL.Image.open(open("/Users/wichert/Desktop/dont.bmp"))
> >>> i
> <PIL.BmpImagePlugin.BmpImageFile instance at 0x43f490>
> >>> i.mode
> 'RGB'
> >>> i.size
> (290, 4294967075L)
> >>> i.tile
> [('raw', (0, 0, 290, 4294967075L), 54, ('BGR', 872, -1))]
>
> and the second image:
>
> >>> i=PIL.Image.open(open("/Users/wichert/Desktop/dont2.bmp"))
> >>> i.mode
> 'RGB'
> >>> i.size
> (288, 4294967078L)
> >>> i.tile
> [('raw', (0, 0, 288, 4294967078L), 54, ('BGR', 864, -1))]
>
>
> Regards,
> Wichert.
>
>
> On 3/5/09 6:41 PM, Fredrik Lundh wrote:
>
>> Can you print the following attributes for a broken file, before calling
>> load:
>>
>>     pil_data.mode
>>     pil_data.size
>>     pil_data.tile
>>
>> </F>
>>
>> 2009/3/3 Wichert Akkerman<wichert at simplon.biz>:
>>
>>
>>> I have a small routine which tries to verify if an uploaded file is a
>>> valid
>>> image. It works like this:
>>>
>>>         image_data=StringIO(data)
>>>         try:
>>>             pil_data=PIL.Image.open(image_data)
>>>         except IOError: # PIL abuses IOError to report parsing errors
>>>             raise Invalid(self.message("bad_image", state),
>>>                     value, state)
>>>
>>>         pil_data.load()
>>>         (width,height)=pil_data.size
>>>
>>> For some images I get an OverflowError on the load() call:
>>>
>>> Module PIL.ImageFile:155 in load
>>>
>>>
>>>>   self.load_prepare()
>>>>>
>>>>>
>>>> Module PIL.ImageFile:223 in load_prepare
>>>
>>>
>>>>   self.im = Image.core.new(self.mode, self.size)
>>>>>
>>>>>
>>>> OverflowError: long int too large to convert to int
>>>
>>> I can not find any information as to what the problem might be. It seems
>>> to
>>> only occur for bmp files; jpeg and png appear to work correctly.
>>>
>>> This is using PIL 1.1.6 and Python 2.5 on a Linux system.
>>>
>>> Wichert.
>>> <<                       self.map = None
>>>
>>>                self.load_prepare()
>>>
>>>                # look for read/seek overrides
>>> <<           if not self.im or\
>>>                   self.im.mode != self.mode or self.im.size != self.size:
>>>                    self.im = Image.core.new(self.mode, self.size)
>>>                # create palette (optional)
>>>                if self.mode == "P":
>>> _______________________________________________
>>> Image-SIG maillist  -  Image-SIG at python.org
>>> http://mail.python.org/mailman/listinfo/image-sig
>>>
>>>
>>>
>>>
>>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>


-- 
Howard Lightstone
hlightstone at gmail.com
howard at eegsoftware.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090306/bd1b499f/attachment.htm>

From fredrik at pythonware.com  Fri Mar  6 18:35:44 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 6 Mar 2009 18:35:44 +0100
Subject: [Image-SIG] towards 1.1.7 (code name "raclette")
Message-ID: <368a5cd50903060935x2b4a269cy816feb2cae06a2e4@mail.gmail.com>

Looks like some people out there have noticed this little page:

    http://bitbucket.org/effbot/pil-2009-raclette/overview/

so I might as well announce it to the list - it's finally time for the
1.1.7 "patch catchup" I mentioned last year.  The plan is simple: pick
must-have/low-hanging/obviously-correct fixes and patches, add them
all on top of 1.1.6, and cut a new 1.1.7 release.  No big overhauls,
no CADT-style rewrites, just plain old boring software maintenance.

The plan is to collect patches/fixes over the next few weeks, and cut
an alpha before the end of the month.

Here's what you can do:

1. Go to the above page and look for your favourite patches.  If it's
there, you don't need to do anything now - just keep an eye on this
list and help us with testing when this thing goes alpha.

2. If the patch or fix isn't there, you can:

   a) followup on this list - either post a patch or a pointer to
      a patch if you have it, or describe the problem.

   b) add it to the issue tracker on that site.

   c) grab a source snapshot (use the 'download >>' feature) and
      prepare a patch; submit it using approach (a) or (b).

   d) or, if you're a Mercurial user, clone the repository and work
      on a patch, and either submit it as as above, send me a pull
      request, or send me a patchbomb (if you don't know what that
      is, please don't ;-).

   e) for sensitive or bulky stuff (e.g. if the test data needed
      to trigger the bug cannot be posted in public), feel free
      to contact me off-list.

As usual, I reserve the right to reject/postpone patches that are too
big or complicated or risk breaking old code.  I also reserve the
right to remove the above repository when this catchup is done; the
final version will be released through the usual channels.

Cheers /F

From fredrik at pythonware.com  Fri Mar  6 18:40:28 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 6 Mar 2009 18:40:28 +0100
Subject: [Image-SIG] towards 1.1.7 (code name "raclette")
In-Reply-To: <368a5cd50903060935x2b4a269cy816feb2cae06a2e4@mail.gmail.com>
References: <368a5cd50903060935x2b4a269cy816feb2cae06a2e4@mail.gmail.com>
Message-ID: <368a5cd50903060940ha2e9d53wa035db32330419c0@mail.gmail.com>

> so I might as well announce it to the list - it's finally time for the
> 1.1.7 "patch catchup" I mentioned last year.

(and many thanks to Tim Hatch who happened to nag me about this at a
point where I had some cycles to spare and happened to be tinkering
with Mercurial for another reason)

</F>

From karsten.hiddemann at mathematik.uni-dortmund.de  Fri Mar  6 18:33:36 2009
From: karsten.hiddemann at mathematik.uni-dortmund.de (Karsten Hiddemann)
Date: Fri, 06 Mar 2009 18:33:36 +0100
Subject: [Image-SIG] towards 1.1.7 (code name "raclette")
In-Reply-To: <368a5cd50903060940ha2e9d53wa035db32330419c0@mail.gmail.com>
References: <368a5cd50903060935x2b4a269cy816feb2cae06a2e4@mail.gmail.com>
	<368a5cd50903060940ha2e9d53wa035db32330419c0@mail.gmail.com>
Message-ID: <49B15E70.9020009@mathematik.uni-dortmund.de>

Fredrik Lundh schrieb:
>> so I might as well announce it to the list - it's finally time for the
>> 1.1.7 "patch catchup" I mentioned last year.

Ah, wonderful.

> (and many thanks to Tim Hatch who happened to nag me about this at a
> point where I had some cycles to spare and happened to be tinkering
> with Mercurial for another reason)

I'd be nagging you constantly, but I was fearing that this might have
had the opposite effect. ;)

Cheers,
Karsten

From fredrik at pythonware.com  Fri Mar  6 19:52:42 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 6 Mar 2009 19:52:42 +0100
Subject: [Image-SIG] OverflowError on image save
In-Reply-To: <368a5cd50903060843y1dfbdb07u89a935941b5e417@mail.gmail.com>
References: <49AD9F41.2050604@simplon.biz>
	<368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>
	<49B12A13.2080406@simplon.biz>
	<29bf24a80903060752h31c2fdd4y6f3114539a178ad@mail.gmail.com>
	<368a5cd50903060843y1dfbdb07u89a935941b5e417@mail.gmail.com>
Message-ID: <368a5cd50903061052m47a33e63s65f01eccdc4fe0cb@mail.gmail.com>

Ok, a patch for this issue can be found here:

    http://bitbucket.org/effbot/pil-2009-raclette/changeset/53b7aa435b9c/

(you can also grab the updated files via that page)

Thanks to Wichert, who reported the problem and provided overflowing samples.

</F>

On Fri, Mar 6, 2009 at 5:43 PM, Fredrik Lundh <fredrik at pythonware.com> wrote:

> Does anyone have a (preferrably) small example, or a simple way to
> create a broken file on a stock Windows XP machine? ?An URL or an
> off-list mail works fine.
>
> </F>

From tim at timhatch.com  Fri Mar  6 21:49:08 2009
From: tim at timhatch.com (Tim Hatch)
Date: Fri, 6 Mar 2009 12:49:08 -0800
Subject: [Image-SIG] PIL and embedded ICC profiles
Message-ID: <576DA127-C143-4299-B9C6-E8EA0D3A1C7C@timhatch.com>

Hiya,

This patch came through last year but it didn't get much review.  I  
was wondering if it's a candidate for PIL 1.1.7 ? I've tested that it  
does indeed pass through profiles in jpeg-to-jpeg and jpeg-to-png  
conversion.

http://mail.python.org/pipermail/image-sig/2008-June/005026.html

There's a version at http://timhatch.com/ex/2009/03/pil-keep-icc-profiles.diff 
  that applies cleanly on raclette edc72df13132 with some whitespace  
cleanups.  I took out most of the unrelated patches, since they'd  
either been fixed already, or (in the case of TIFF resolution  
information) I'm not familiar with the code and it didn't apply  
cleanly.  I suppose that's a separate topic, and I'm missing a pointer  
to the original issue.

Tim

From fredrik at pythonware.com  Fri Mar  6 21:49:45 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 6 Mar 2009 21:49:45 +0100
Subject: [Image-SIG] [raclette] help wanted - anyone has a recent photoshop
	within reach and a few minutes to spare?
Message-ID: <368a5cd50903061249h243d3b57qbc2a9998da6c7b9b@mail.gmail.com>

I'm have a lingering feeling that someone have helped me with this
before, but I cannot find the mails/files, so here we go again:

I need a small (e.g. 100x100) sample CMYK JPEG from Photoshop - that
is, the kind of file that keeps causing trouble for people on this
list from time to time.  To make it easy for me to work with this
file, do something like this:

1. create a small CMYK image, and fill it with solid black or solid white
2. pick a nice font and draw the text "RED" in big red letters
3. put a small red object in the upper left corner (make sure the
corner pixel and at least a few pixels around it are red)
4. save as JPEG.
5. convert to RGB, and save another JPEG copy of the same image.
6. mail both images to me and/or the list

Thanks /F

From fredrik at pythonware.com  Fri Mar  6 21:58:32 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 6 Mar 2009 21:58:32 +0100
Subject: [Image-SIG] PIL and embedded ICC profiles
In-Reply-To: <576DA127-C143-4299-B9C6-E8EA0D3A1C7C@timhatch.com>
References: <576DA127-C143-4299-B9C6-E8EA0D3A1C7C@timhatch.com>
Message-ID: <368a5cd50903061258k283ddc55jd95e3da2055a6653@mail.gmail.com>

Except for the JPEG hack, Florian's original patch looks pretty quite
clean to me.  And I'm pretty sure I can do something about the hack
before we go alpha.

That diff link looks a bit strange to me, though - looks more like
status output than diff output.  Am I missing something, or did you
post the wrong file?  I could pull the patch from your Hg repository
if that's easier for you.

</F>

On Fri, Mar 6, 2009 at 9:49 PM, Tim Hatch <tim at timhatch.com> wrote:
> Hiya,
>
> This patch came through last year but it didn't get much review. ?I was
> wondering if it's a candidate for PIL 1.1.7 ? I've tested that it does
> indeed pass through profiles in jpeg-to-jpeg and jpeg-to-png conversion.
>
> http://mail.python.org/pipermail/image-sig/2008-June/005026.html
>
> There's a version at
> http://timhatch.com/ex/2009/03/pil-keep-icc-profiles.diff?that applies
> cleanly on raclette edc72df13132 with some whitespace cleanups. ?I took out
> most of the unrelated patches, since they'd either been fixed already, or
> (in the case of TIFF resolution information) I'm not familiar with the code
> and it didn't apply cleanly. ?I suppose that's a separate topic, and I'm
> missing a pointer to the original issue.
>
> Tim
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From fredrik at pythonware.com  Fri Mar  6 22:11:20 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 6 Mar 2009 22:11:20 +0100
Subject: [Image-SIG] [raclette] help wanted - anyone has a recent
	photoshop within reach and a few minutes to spare?
In-Reply-To: <368a5cd50903061249h243d3b57qbc2a9998da6c7b9b@mail.gmail.com>
References: <368a5cd50903061249h243d3b57qbc2a9998da6c7b9b@mail.gmail.com>
Message-ID: <368a5cd50903061311p3d51a2adh948a8012439af08f@mail.gmail.com>

Problem solved!

Roughly half of the display tools I have on my machine displays it as
blue on gray, and the rest as red on black.  This is going to be
rather interesting ;-)

</F>

On Fri, Mar 6, 2009 at 9:49 PM, Fredrik Lundh <fredrik at pythonware.com> wrote:
> I'm have a lingering feeling that someone have helped me with this
> before, but I cannot find the mails/files, so here we go again:
>
> I need a small (e.g. 100x100) sample CMYK JPEG from Photoshop - that
> is, the kind of file that keeps causing trouble for people on this
> list from time to time. ?To make it easy for me to work with this
> file, do something like this:
>
> 1. create a small CMYK image, and fill it with solid black or solid white
> 2. pick a nice font and draw the text "RED" in big red letters
> 3. put a small red object in the upper left corner (make sure the
> corner pixel and at least a few pixels around it are red)
> 4. save as JPEG.
> 5. convert to RGB, and save another JPEG copy of the same image.
> 6. mail both images to me and/or the list
>
> Thanks /F
>

From charlie at begeistert.org  Fri Mar  6 22:16:13 2009
From: charlie at begeistert.org (Charlie Clark)
Date: Fri, 6 Mar 2009 22:16:13 +0100
Subject: [Image-SIG] [raclette] help wanted - anyone has a recent
	photoshop within reach and a few minutes to spare?
In-Reply-To: <368a5cd50903061249h243d3b57qbc2a9998da6c7b9b@mail.gmail.com>
References: <368a5cd50903061249h243d3b57qbc2a9998da6c7b9b@mail.gmail.com>
Message-ID: <D334AEAB-FC62-4B1B-82B1-66519AEA9580@begeistert.org>


Am 06.03.2009 um 21:49 schrieb Fredrik Lundh:

> I'm have a lingering feeling that someone have helped me with this
> before, but I cannot find the mails/files, so here we go again:
>
> I need a small (e.g. 100x100) sample CMYK JPEG from Photoshop - that
> is, the kind of file that keeps causing trouble for people on this
> list from time to time.  To make it easy for me to work with this
> file, do something like this:
>
> 1. create a small CMYK image, and fill it with solid black or solid  
> white
> 2. pick a nice font and draw the text "RED" in big red letters
> 3. put a small red object in the upper left corner (make sure the
> corner pixel and at least a few pixels around it are red)
> 4. save as JPEG.
> 5. convert to RGB, and save another JPEG copy of the same image.
> 6. mail both images to me and/or the list


'Ere you go.
--
Charlie Clark
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pil_sample_cmyk.jpg
Type: image/jpeg
Size: 29364 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090306/9208cbb5/attachment-0002.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pil_sample_rgb.jpg
Type: image/jpeg
Size: 29342 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090306/9208cbb5/attachment-0003.jpg>
-------------- next part --------------

Helmholtzstr. 20
D?sseldorf
D- 40215
Tel: +49-211-938-5360
GSM: +49-178-782-6226




From tim at timhatch.com  Fri Mar  6 22:25:09 2009
From: tim at timhatch.com (Tim Hatch)
Date: Fri, 6 Mar 2009 13:25:09 -0800
Subject: [Image-SIG] PIL and embedded ICC profiles
In-Reply-To: <368a5cd50903061258k283ddc55jd95e3da2055a6653@mail.gmail.com>
References: <576DA127-C143-4299-B9C6-E8EA0D3A1C7C@timhatch.com>
	<368a5cd50903061258k283ddc55jd95e3da2055a6653@mail.gmail.com>
Message-ID: <84D83908-E583-4AD3-9BBD-95C8933BBA02@timhatch.com>

> Except for the JPEG hack, Florian's original patch looks pretty quite
> clean to me.

Right, I just fixed where hunks didn't apply and adjusted to be more  
PEP8.

> That diff link looks a bit strange to me, though - looks more like
> status output than diff output.  Am I missing something, or did you
> post the wrong file?

You're correct.  The diff itself is up there now.

> http://timhatch.com/ex/2009/03/pil-keep-icc-profiles.diff

Tim


From bbrace at eskimo.com  Fri Mar  6 22:22:36 2009
From: bbrace at eskimo.com ({ brad brace })
Date: Fri, 6 Mar 2009 13:22:36 -0800 (PST)
Subject: [Image-SIG] 12hr project
In-Reply-To: <mailman.5207.1236374231.11745.image-sig@python.org>
References: <mailman.5207.1236374231.11745.image-sig@python.org>
Message-ID: <Pine.SUN.4.58.0903061322130.9477@eskimo.com>


            _  |__   __| |          /_ |__ \| |
                      | __|   | | | (_) | |  __/ (__| |_
               __ | |  | | | |  __/  | |/ /_| | | | |
                   _  | |  | '_ \ / _ \  | | / /| '_ \| '__|

The 12hr-ISBN-JPEG Project       >>>> posted since 1994 <<<<

           _  | |  | '_ \ / _ \  | | / /| '_ \| '__|
                 -_    | |  | |__   ___   | |  ) | |__  _ __
           _ |  __ \         (_)          | |

"An impassive observer, { brad brace } forges a personal aesthetic in
these 12hr-images infused with blank-sadness and a sense of mystery.
What makes them both new and significant is the fact that he organizes
its contents in sequences, applying the principles of cinematographic
montage to fixed images."


 You begin to sense the byshadows that stretch from the awe of global
 dominance. How the intersecting systems help pull us apart, leaving
 us vague, drained, docile, soft in our inner discourse, willing to be
 shaped, to be overwhelmed -- easy retreats, half beliefs. Works of art
 are complex formal interventions within discursive traditions and their
 myriad filiations. These interventions are defined precisely by their
 incomparable capacity to trace the dynamics of historical process in
 paradoxical gestures of simultaneously prognostic and mnemonic
 temporalities.
                      | __|   | | | (_) | |  __/ (__| |_
              _  | |  | '_ \ / _ \  | | / /| '_ \| '__|
                   _| |__) | __ ___  _  ___  ___| |_
                 |_  ___/ '__/ _ \| |/ _ \/ __| __|
              |_| _  |_|  \___/| |\___|\___|\__|
                           _          _/ |
                             _        |__/


> > > > Synopsis: The 12hr-ISBN-JPEG Project began December 30, 1994. A
`round-the-clock posting of sequenced hypermodern imagery from { brad
brace }. The hypermodern minimizes the familiar, the known, the
recognizable; it suspends identity, relations and history. This
discourse, far from determining the locus in which it speaks, is avoiding
the ground on which it could find support. It is trying to operate a
decentering that leaves no privilege to any center.


                       The 12-hour ISBN JPEG Project
                       -----------------------------
                          began December 30, 1994


  Pointless Hypermodern Imagery... posted/mailed every 12 hours... a
spectral, trajective alignment for the 00`s! A continuum of minimalist
masks in the face of catastrophe; conjuring up transformative metaphors
for the everyday... A poetic reversibility of exclusive events...

        A post-rhetorical, continuous, apparently random sequence of
imagery...  genuine gritty, greyscale...  corruptable, compact,
collectable and compelling convergence. The voluptuousness of the grey
imminence: the art of making the other disappear. Continual visual
impact; an optical drumming, sculpted in duration, on the endless present
of the Net.

  An extension of the printed ISBN-Book (0-9690745) series... critically
unassimilable... imagery is gradually acquired, selected and re-sequenced
over time...  ineluctable, vertiginous connections. The 12hr dialtone...

		   [ see http://www.eskimo.com/~bbrace/netcom/books.txt ]

KEYWORDS:

>> Disconnected, disjunctive, distended, de-centered,
   de-composed, ambiguous, augmented, ambilavent, homogeneous, reckless...
>> Multi-faceted, oblique, obsessive, obscure, obdurate...
>> Promulgated, personal, permeable, prolonged, polymorphous, provocative,
   poetic, plural, perverse, potent, prophetic, pathological, pointless,
   placeholders...
>> Robust, real, redundant, resplendent, revolutionary, redeeming...
>> Emergent, evolving, eccentric, eclectic, egregious, eternal, exciting,
   entertaining, evasive, entropic, erotic, entrancing, enduring, ergodic,
   expansive...


        Every 12 hours, another!...  view them, re-post `em, save `em,
trade `em, print `em, even publish them...

Here`s how:

~ Set www-links to ->  http://www.eskimo.com/~bbrace/12hr.html
                   ->  http://bbrace.laughingsquid.net/12hr.html
                   ->  http://bbrace.net/12hr.html
		   ->  http://noemata.net/12hr/

  Look for the 12-hr-icon. Heavy traffic may require you to specify files
  more than once! Anarchie, Fetch, CuteFTP, TurboGopher...

~ Download from ->  ftp.rdrop.com   /pub/users/bbrace
  Download from ->  ftp.eskimo.com  /u/b/bbrace
  Download from ->  hotline://artlyin.ftr.va.com.au
  Download from ->  http://kunst.noemata.net/12hr/

  * Remember to set tenex or binary. Get 12hr.jpeg

~ E-mail -> If you only have access to email, then you can use FTPmail to
  do essentially the same thing. Send a message with a body of 'help' to
  the server address nearest you:
  *
  ftpmail at ccc.uba.ar                    ftpmail at cs.uow.edu.au
  ftpmail at ftp.uni-stuttgart.de          ftpmail at ftp.Dartmouth.edu
  ftpmail at ieunet.ie                     ftpmail at src.doc.ic.ac.uk
  ftpmail at archie.inesc.pt               ftpmail at ftp.sun.ac.za
  ftpmail at ftp.sunet.se                  ftpmail at ftp.luth.se
  ftpmail at NCTUCCCA.edu.tw               ftpmail at oak.oakland.edu
  ftpmail at sunsite.unc.edu               ftpmail at decwrl.dec.com
  ftpmail at census.gov
  bitftp at plearn.bitnet                  bitftp at dearn.bitnet
  bitftp at vm.gmd.de                      bitftp at plearn.edu.pl
  bitftp at pucc.princeton.edu             bitftp at pucc.bitnet
  *                                     *


~ Mirror-sites requested! Archives too!
  The latest new jpeg will always be named, 12hr.jpeg
  Average size of images is only 45K.
  *
  Perl program to mirror ftp-sites/sub-directories:
  src.doc.ic.ac.uk:/packages/mirror
  *

~ Postings to usenet newsgroups:
  12hr
  alt.12hr
  alt.binaries.pictures.12hr
  alt.binaries.pictures.misc
  alt.binaries.pictures.fine-art.misc

* * Ask your system's news-administrator to carry these groups!
  (There are also usenet image browsers: TIFNY, PluckIt, Picture Agent,
    PictureView, Extractor97, NewsRover, Binary News Assistant, EasyNews)

~ This interminable, relentless (online) sequence of imagery began in
earnest on December 30, 1994. The basic structure of the project has been
over twenty-five years in the making. While the specific sequence of
photographs has been presently orchestrated for many years` worth of
12-hour postings, I will undoubtedly be tempted to tweak the ongoing
publication with additional new interjected imagery. Each 12-hour image
is like the turning of a page; providing ample time for reflection,
interruption, and assimilation.

~ The sites listed above also contain information on other cultural
projects and sources.

~ A very low-volume, moderated mailing list for announcements and
occasional commentary related to this project has been established at
topica.com /subscribe 12hr-isbn-jpeg

 --
 	 The image was to make nothing visible but their connection with
 	 one another by space and air, yet each surrounded by the unique
 	 aura that disengages every deeply seen image from the world of
 	 irrelevant relationships and calls forth a tremor of
 	 astonishment at its fateful necessity. Thus from artworks of
 	 dead masters, over-life-size strangeness whose names we do not
 	 know and do not wish to know, look out at us enigmatically as
 	 symbols of all being.

 --

 Big Grey Bricks: This project also serves as a rehearsal for its
 culmination as a series of offset-printed volumes: each 800+ full-bleed
 pages (5x8"_300lpi), where the full integrated rhythm of
 greyscale-sequence can be more intricately resolved. I'd provide all
 design, prepress and production.

 --

This project remains untainted by corrupt corporate and glib government
art-subsidies. Some opportunities still exist for financially assisting the
publication of editions of large (33x46") prints; perhaps (Iris giclees)
inkjet duotones or extended-black quadtones.

Other supporters receive rare copies of the first three web-offset printed
ISBN-Books. Contributions and requests for 12hr-email-subscriptions,
can also be made at http://bbrace.laughingsquid.net/buy-into.html,
or by mailed cheque/check: $5/mo $50/yr. Art-institutions must pay $12K
for each image retained longer than 12 hours.

--

ISBN is International Standard Book Number. JPEG and GIF are types of
image files. Get the text-file, 'pictures-faq' to learn how to view or
translate these images. [http://www.eskimo.com/~bbrace/netcom/pictures
-faq.html]

-- 

(c) Credit appreciated. Copyleft

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009


<bbrace at eskimo.com>





From fredrik at pythonware.com  Fri Mar  6 22:52:29 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 6 Mar 2009 22:52:29 +0100
Subject: [Image-SIG] PIL and embedded ICC profiles
In-Reply-To: <84D83908-E583-4AD3-9BBD-95C8933BBA02@timhatch.com>
References: <576DA127-C143-4299-B9C6-E8EA0D3A1C7C@timhatch.com>
	<368a5cd50903061258k283ddc55jd95e3da2055a6653@mail.gmail.com>
	<84D83908-E583-4AD3-9BBD-95C8933BBA02@timhatch.com>
Message-ID: <368a5cd50903061352y7abeddb3h6ab97a1628335f2f@mail.gmail.com>

Thanks!

http://bitbucket.org/effbot/pil-2009-raclette/changeset/debdde9b7862/

</F>

On Fri, Mar 6, 2009 at 10:25 PM, Tim Hatch <tim at timhatch.com> wrote:
>> Except for the JPEG hack, Florian's original patch looks pretty quite
>> clean to me.
>
> Right, I just fixed where hunks didn't apply and adjusted to be more PEP8.
>
>> That diff link looks a bit strange to me, though - looks more like
>> status output than diff output. ?Am I missing something, or did you
>> post the wrong file?
>
> You're correct. ?The diff itself is up there now.
>
>> http://timhatch.com/ex/2009/03/pil-keep-icc-profiles.diff
>
> Tim
>
>

From fredrik at pythonware.com  Sat Mar  7 01:26:15 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 7 Mar 2009 01:26:15 +0100
Subject: [Image-SIG] [raclette] that JPEG CMYK thing, again
Message-ID: <368a5cd50903061626y68cd64f9u1615b8ab9564081a@mail.gmail.com>

Thanks to Tim and Charlie, I now have a set of nice Photoshop CMYK
samples to play with.  The first result is a revamp of Kevin's CMYK
patch; I'm now doing the inversion at the codec levels, both when
reading and writing.  Also, the code *always* assumes Adobe behaviour,
in both directions.  The current version of the plugin is here:

    http://bitbucket.org/effbot/pil-2009-raclette/src/e351bc0228c4/PIL/JpegImagePlugin.py

I *think* this does the right thing, but I'm not 100% sure.  In other
words, it would be helpful if those of you who've struggled with CMYK
stuff can take this version for a test drive; it should work just fine
together with a standard 1.1.6.

</F>

From wichert at simplon.biz  Sun Mar  8 18:09:13 2009
From: wichert at simplon.biz (Wichert Akkerman)
Date: Sun, 08 Mar 2009 18:09:13 +0100
Subject: [Image-SIG] OverflowError on image save
In-Reply-To: <368a5cd50903060843y1dfbdb07u89a935941b5e417@mail.gmail.com>
References: <49AD9F41.2050604@simplon.biz>	
	<368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>	
	<49B12A13.2080406@simplon.biz>	
	<29bf24a80903060752h31c2fdd4y6f3114539a178ad@mail.gmail.com>
	<368a5cd50903060843y1dfbdb07u89a935941b5e417@mail.gmail.com>
Message-ID: <49B3FBB9.3090204@simplon.biz>

I sent two images demonstrating the problem to Fredrik off-list.

I suspect this is a fairly common problem: I have seen this error occur 
several when users try to upload a member portrait to a community site, 
and at this moment only a handful users have access to it since it is 
still in a closed test phase.

Wichert.



On 3/6/09 5:43 PM, Fredrik Lundh wrote:
> Does anyone have a (preferrably) small example, or a simple way to
> create a broken file on a stock Windows XP machine?  An URL or an
> off-list mail works fine.
>
> </F>
>
> On Fri, Mar 6, 2009 at 4:52 PM, Howard Lightstone
> <howard at eegsoftware.com>  wrote:
>    
>> This LOOKS like a bug in Windows XP which creates .BMP files via DirectX
>> with an incorrect bitmap header size (54 vice 40).  Changing the 54 to 40
>> makes the file readable again.  Most utilities seem to ignore this but
>> Windows itself won't be able to 'read' it as a .BMP file.
>>
>> (using XP SP2 and DirectX 9.0c ... I haven't tried this on Vista yet)
>>
>> On Fri, Mar 6, 2009 at 5:50 AM, Wichert Akkerman<wichert at simplon.biz>
>> wrote:
>>      
>>> The first image:
>>>
>>>        
>>>>>> i=PIL.Image.open(open("/Users/wichert/Desktop/dont.bmp"))
>>>>>> i
>>>>>>              
>>> <PIL.BmpImagePlugin.BmpImageFile instance at 0x43f490>
>>>        
>>>>>> i.mode
>>>>>>              
>>> 'RGB'
>>>        
>>>>>> i.size
>>>>>>              
>>> (290, 4294967075L)
>>>        
>>>>>> i.tile
>>>>>>              
>>> [('raw', (0, 0, 290, 4294967075L), 54, ('BGR', 872, -1))]
>>>
>>> and the second image:
>>>
>>>        
>>>>>> i=PIL.Image.open(open("/Users/wichert/Desktop/dont2.bmp"))
>>>>>> i.mode
>>>>>>              
>>> 'RGB'
>>>        
>>>>>> i.size
>>>>>>              
>>> (288, 4294967078L)
>>>        
>>>>>> i.tile
>>>>>>              
>>> [('raw', (0, 0, 288, 4294967078L), 54, ('BGR', 864, -1))]
>>>
>>>
>>> Regards,
>>> Wichert.
>>>
>>> On 3/5/09 6:41 PM, Fredrik Lundh wrote:
>>>        
>>>> Can you print the following attributes for a broken file, before calling
>>>> load:
>>>>
>>>>      pil_data.mode
>>>>      pil_data.size
>>>>      pil_data.tile
>>>>
>>>> </F>
>>>>
>>>> 2009/3/3 Wichert Akkerman<wichert at simplon.biz>:
>>>>
>>>>          
>>>>> I have a small routine which tries to verify if an uploaded file is a
>>>>> valid
>>>>> image. It works like this:
>>>>>
>>>>>          image_data=StringIO(data)
>>>>>          try:
>>>>>              pil_data=PIL.Image.open(image_data)
>>>>>          except IOError: # PIL abuses IOError to report parsing errors
>>>>>              raise Invalid(self.message("bad_image", state),
>>>>>                      value, state)
>>>>>
>>>>>          pil_data.load()
>>>>>          (width,height)=pil_data.size
>>>>>
>>>>> For some images I get an OverflowError on the load() call:
>>>>>
>>>>> Module PIL.ImageFile:155 in load
>>>>>
>>>>>            
>>>>>>>    self.load_prepare()
>>>>>>>
>>>>>>>                
>>>>> Module PIL.ImageFile:223 in load_prepare
>>>>>
>>>>>            
>>>>>>>    self.im = Image.core.new(self.mode, self.size)
>>>>>>>
>>>>>>>                
>>>>> OverflowError: long int too large to convert to int
>>>>>
>>>>> I can not find any information as to what the problem might be. It seems
>>>>> to
>>>>> only occur for bmp files; jpeg and png appear to work correctly.
>>>>>
>>>>> This is using PIL 1.1.6 and Python 2.5 on a Linux system.
>>>>>
>>>>> Wichert.
>>>>> <<                         self.map = None
>>>>>
>>>>>                 self.load_prepare()
>>>>>
>>>>>                 # look for read/seek overrides
>>>>> <<             if not self.im or\
>>>>>                    self.im.mode != self.mode or self.im.size !=
>>>>> self.size:
>>>>>                     self.im = Image.core.new(self.mode, self.size)
>>>>>                 # create palette (optional)
>>>>>                 if self.mode == "P":
>>>>> _______________________________________________
>>>>> Image-SIG maillist  -  Image-SIG at python.org
>>>>> http://mail.python.org/mailman/listinfo/image-sig
>>>>>
>>>>>
>>>>>
>>>>>            
>>> _______________________________________________
>>> Image-SIG maillist  -  Image-SIG at python.org
>>> http://mail.python.org/mailman/listinfo/image-sig
>>>
>>>        
>>
>> --
>> Howard Lightstone
>> hlightstone at gmail.com
>> howard at eegsoftware.com
>>
>>      


From charlie at begeistert.org  Sun Mar  8 19:05:09 2009
From: charlie at begeistert.org (Charlie Clark)
Date: Sun, 8 Mar 2009 19:05:09 +0100
Subject: [Image-SIG] OverflowError on image save
In-Reply-To: <49B3FBB9.3090204@simplon.biz>
References: <49AD9F41.2050604@simplon.biz>	
	<368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>	
	<49B12A13.2080406@simplon.biz>	
	<29bf24a80903060752h31c2fdd4y6f3114539a178ad@mail.gmail.com>
	<368a5cd50903060843y1dfbdb07u89a935941b5e417@mail.gmail.com>
	<49B3FBB9.3090204@simplon.biz>
Message-ID: <C4AEB55E-ED47-4BA1-BA15-65C1105615C7@begeistert.org>


Am 08.03.2009 um 18:09 schrieb Wichert Akkerman:

> I sent two images demonstrating the problem to Fredrik off-list.
>
> I suspect this is a fairly common problem: I have seen this error  
> occur several when users try to upload a member portrait to a  
> community site, and at this moment only a handful users have access  
> to it since it is still in a closed test phase.


Hiya Wichert,

I'm just curious as to why a .BMP is likely to be a common format - is  
the image coming as .BMP from some common windows control?

Charlie
--
Charlie Clark
Helmholtzstr. 20
D?sseldorf
D- 40215
Tel: +49-211-938-5360
GSM: +49-178-782-6226




From wichert at simplon.biz  Sun Mar  8 19:09:53 2009
From: wichert at simplon.biz (Wichert Akkerman)
Date: Sun, 08 Mar 2009 19:09:53 +0100
Subject: [Image-SIG] OverflowError on image save
In-Reply-To: <C4AEB55E-ED47-4BA1-BA15-65C1105615C7@begeistert.org>
References: <49AD9F41.2050604@simplon.biz>	
	<368a5cd50903050941h2960a6c6j28eb8dbbc08866fb@mail.gmail.com>	
	<49B12A13.2080406@simplon.biz>	
	<29bf24a80903060752h31c2fdd4y6f3114539a178ad@mail.gmail.com>
	<368a5cd50903060843y1dfbdb07u89a935941b5e417@mail.gmail.com>
	<49B3FBB9.3090204@simplon.biz>
	<C4AEB55E-ED47-4BA1-BA15-65C1105615C7@begeistert.org>
Message-ID: <49B409F1.8080704@simplon.biz>

Hi Charlie,

On 3/8/09 7:05 PM, Charlie Clark wrote:
> Hiya Wichert,
>
> I'm just curious as to why a .BMP is likely to be a common format - is 
> the image coming as .BMP from some common windows control?

I have no idea. I just noticed that we are currently testing a new site 
with 50 users, and two have hit this problem. Personally I'm shocked 
people still use BMPs,

Wichert.

From fredrik at pythonware.com  Mon Mar  9 18:23:12 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 9 Mar 2009 18:23:12 +0100
Subject: [Image-SIG] [raclette] PyCMS now part of PIL
Message-ID: <368a5cd50903091023q36faa3dep28b3c66f6635cedd@mail.gmail.com>

Kevin Cazabon has graciously donated his PyCMS library to PIL, and it
landed in the "raclette" tree yesterday.  The library will
automatically be built if you have LittleCMS on the machine
(www.littlecms.com).  The Python binding is now in PIL.ImageCms.

I'm still working on a tighter integration, but the module provides
the procedural PyCMS API (with a few enhancements; you can e.g. read
profiles from memory by passing in file objects instead of filenames).
 In other words, code written for pyCMS "should" still work if you
just change "import pyCMS" to

    import PIL.ImageCms as pyCMS

The internals are reworked, though, so code that rely on
implementation details, the exact phrasing of error messages (the PIL
version is less informative), or access to the pyCMSdll layer won't
work.

If you want to help out testing the integration, you can pull the
current snapshot from

    http://bitbucket.org/effbot/pil-2009-raclette/overview/

(either as a zip/tarball via the download link, or using "hg clone").

Cheers /F

From kevin at cazabon.com  Tue Mar 10 00:32:10 2009
From: kevin at cazabon.com (Kevin Cazabon)
Date: Mon, 9 Mar 2009 19:32:10 -0400
Subject: [Image-SIG] [raclette] PyCMS now part of PIL
In-Reply-To: <368a5cd50903091023q36faa3dep28b3c66f6635cedd@mail.gmail.com>
References: <368a5cd50903091023q36faa3dep28b3c66f6635cedd@mail.gmail.com>
Message-ID: <4CA150B4-23CB-4BA9-A4F2-F0F8E282BA37@cazabon.com>

Thanks for the hard work on upgrading and integrating this Fredrik -  
I've been following your changes and it seems like quite a significant  
amount of work on your end to clean it up and streamline things.  I'm  
looking forward to playing with the in-memory profile reading, along  
with the embedded profile reading that other SIG members have been  
working on.  Having CMS in PIL will be yet another formidable tool for  
PIL and adoption in high-end applications.

In any case, I'm proud to contribute pyCMS - although it's really just  
a wrapper around Marti Maria's harder work on the LittleCMS library  
itself.

Kevin.


On 09 Mar 2009, at 13:23, Fredrik Lundh wrote:

> Kevin Cazabon has graciously donated his PyCMS library to PIL, and it
> landed in the "raclette" tree yesterday.  The library will
> automatically be built if you have LittleCMS on the machine
> (www.littlecms.com).  The Python binding is now in PIL.ImageCms.
>
> I'm still working on a tighter integration, but the module provides
> the procedural PyCMS API (with a few enhancements; you can e.g. read
> profiles from memory by passing in file objects instead of filenames).
> In other words, code written for pyCMS "should" still work if you
> just change "import pyCMS" to
>
>    import PIL.ImageCms as pyCMS
>
> The internals are reworked, though, so code that rely on
> implementation details, the exact phrasing of error messages (the PIL
> version is less informative), or access to the pyCMSdll layer won't
> work.
>
> If you want to help out testing the integration, you can pull the
> current snapshot from
>
>    http://bitbucket.org/effbot/pil-2009-raclette/overview/
>
> (either as a zip/tarball via the download link, or using "hg clone").
>
> Cheers /F
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig


From nikolai at netlife.no  Tue Mar 10 09:09:30 2009
From: nikolai at netlife.no (Nikolai Ugelvik)
Date: Tue, 10 Mar 2009 09:09:30 +0100
Subject: [Image-SIG] XVThumbImagePlugin infinite loop fix
Message-ID: <49B6203A.9080509@netlife.no>

I sent an email regarding a bug with this module, including a patch. 
Would like to have this included in 1.1.7.

-Nikolai


Original message:

Hi.

The PIL module XVThumbImagePlugin.py has an issue with XVThumb files 
that do not include #END_OF_COMMENTS (or a comment section at all). I'm 
not quite sure if the problem is an invalid file (is the comment 
required?), but we encountered a file like this. Anyway, the result is 
XVThumbImagePlugin being stuck in an infinite loop. The patch supports 
files without comments and/or comment sections without #END_OF_COMMENTS.

Attached is a diff against XVThumbImagePlugin.py from PIL version 1.1.6. 
(I didn't figure out where else to send the diff.)

-Nikolai
-------------- next part --------------
A non-text attachment was scrubbed...
Name: XVThumbImagePlugin.diff
Type: text/x-patch
Size: 789 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090310/e9d6c87a/attachment.bin>

From lists+Image_SIG at hoech.org  Tue Mar 10 13:45:20 2009
From: lists+Image_SIG at hoech.org (=?ISO-8859-1?Q?Florian_H=F6ch?=)
Date: Tue, 10 Mar 2009 13:45:20 +0100
Subject: [Image-SIG] [raclette] PyCMS now part of PIL
In-Reply-To: <368a5cd50903091023q36faa3dep28b3c66f6635cedd@mail.gmail.com>
References: <368a5cd50903091023q36faa3dep28b3c66f6635cedd@mail.gmail.com>
Message-ID: <49B660E0.5050400@hoech.org>

Very nice - it just needs LCMS flags support so we can use black point
compensation and softproofing :) See attached patches. They also fix a
naming issue which could cause confusion - display profile (ImageCms
wording) actually means proof profile (lcms wording), so I changed
variable names and docstrings where applicable. Patches are tested under
Python 2.6.

Regards,

Florian H?ch

Fredrik Lundh schrieb:
> Kevin Cazabon has graciously donated his PyCMS library to PIL, and it
> landed in the "raclette" tree yesterday.  The library will
> automatically be built if you have LittleCMS on the machine
> (www.littlecms.com).  The Python binding is now in PIL.ImageCms.
> 
> I'm still working on a tighter integration, but the module provides
> the procedural PyCMS API (with a few enhancements; you can e.g. read
> profiles from memory by passing in file objects instead of filenames).
>  In other words, code written for pyCMS "should" still work if you
> just change "import pyCMS" to
> 
>     import PIL.ImageCms as pyCMS
> 
> The internals are reworked, though, so code that rely on
> implementation details, the exact phrasing of error messages (the PIL
> version is less informative), or access to the pyCMSdll layer won't
> work.
> 
> If you want to help out testing the integration, you can pull the
> current snapshot from
> 
>     http://bitbucket.org/effbot/pil-2009-raclette/overview/
> 
> (either as a zip/tarball via the download link, or using "hg clone").
> 
> Cheers /F
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: _imagingcms.c.diff
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090310/c333765b/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: ImageCms.py.diff
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090310/c333765b/attachment-0001.txt>

From oliver at weichhold.com  Thu Mar 12 22:30:47 2009
From: oliver at weichhold.com (Oliver Weichhold)
Date: Thu, 12 Mar 2009 22:30:47 +0100
Subject: [Image-SIG] Strange Truetype ImageFont rendering behavior
Message-ID: <a13da490903121430m5e5856a8rf5a3e01745b67e06@mail.gmail.com>

The code snippet below produces the following two result images on my
system:
Arial:
http://farm4.static.flickr.com/3658/3349324835_ecbb460afb_o.jpg

TrajanPro-Bold:
http://farm4.static.flickr.com/3439/3349324795_71b1c1fd48_o.jpg

I am totally puzzled as to why the font size is kind of applied horizontally
for Trajan Bold. I've produced the images on a freshly installed Windows XP
64 with Python 2.61 + PIL 1.1.6.

Any suggestions what's going on here?

<snip>
from PIL import Image, ImageOps
from PIL import ImageFont, ImageDraw

arial = ImageFont.truetype("Arial.ttf", 24)
trajan = ImageFont.truetype("TrajanPro-Bold.otf", 24)

def render_msg(font, filename):
im = Image.new("RGB", (200,50), "#ddd")
draw = ImageDraw.Draw(im)
draw.text((10,10), "Run awayyyy!", font=font, fill="red")
im.save(filename)

render_msg(arial, 'f1.jpg')
render_msg(trajan, 'f2.jpg')
</snip>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090312/a48ca156/attachment.htm>

From fredrik at pythonware.com  Fri Mar 13 12:08:10 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 13 Mar 2009 12:08:10 +0100
Subject: [Image-SIG] Strange Truetype ImageFont rendering behavior
In-Reply-To: <a13da490903121430m5e5856a8rf5a3e01745b67e06@mail.gmail.com>
References: <a13da490903121430m5e5856a8rf5a3e01745b67e06@mail.gmail.com>
Message-ID: <368a5cd50903130408x5b296984id99952ef1182f48d@mail.gmail.com>

Seems to work for me (see attachment).  However, I just realized that
I have a rather old version of Freetype on this machine (2.1.10).
What version are you using?

</F>

2009/3/12 Oliver Weichhold <oliver at weichhold.com>:
> The code snippet below produces the following two result images on my
> system:
> Arial:
> http://farm4.static.flickr.com/3658/3349324835_ecbb460afb_o.jpg
>
> TrajanPro-Bold:
> http://farm4.static.flickr.com/3439/3349324795_71b1c1fd48_o.jpg
>
> I am totally puzzled as to why the font size is kind of applied horizontally
> for Trajan Bold. I've produced the images on a freshly installed Windows XP
> 64 with Python 2.61 + PIL 1.1.6.
> Any suggestions what's going on here?
> <snip>
> from PIL import Image, ImageOps
> from PIL import ImageFont, ImageDraw
> arial = ImageFont.truetype("Arial.ttf", 24)
> trajan = ImageFont.truetype("TrajanPro-Bold.otf", 24)
> def render_msg(font, filename):
> im = Image.new("RGB", (200,50), "#ddd")
> draw = ImageDraw.Draw(im)
> draw.text((10,10), "Run awayyyy!", font=font, fill="red")
> im.save(filename)
> render_msg(arial, 'f1.jpg')
> render_msg(trajan, 'f2.jpg')
> </snip>
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.gif
Type: image/gif
Size: 2379 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090313/6705218e/attachment.gif>

From evg.salmin at gmail.com  Fri Mar 13 23:03:52 2009
From: evg.salmin at gmail.com (Evgeny Salmin)
Date: Sat, 14 Mar 2009 04:03:52 +0600
Subject: [Image-SIG] Converting images to palette mode
Message-ID: <d7f793bd0903131503r4378cddfhe1cd9eac517a0bd0@mail.gmail.com>

Hello.

   Recently we wrote about memory problems of PIL while converting large
number of png files to 8-bit palette format. After some number of messages
simple converting script (see in attach) takes all available memory from a
system and produces QuantizationError. After some examinating of PIL source
code we've founded that QuantizationError arrises after failed malloc. Each
new image increase memory usage, whereas this script should work in fixed
amount of memory (it processes image, then takes the next one).
I've realized that problem can be solved by applying simple patch (see
attach). After tests on huge number of png files memory consumption looks
fixed and converting works correct.
I'd like to know PIL author's opinion. Thank you.

Evgeny
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090314/f5f382a0/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PIL_mem_leak.diff
Type: application/octet-stream
Size: 348 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090314/f5f382a0/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: to_8bit.py
Type: application/octet-stream
Size: 2160 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090314/f5f382a0/attachment-0001.obj>

From fredrik at pythonware.com  Sat Mar 14 00:04:34 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 14 Mar 2009 00:04:34 +0100
Subject: [Image-SIG] Converting images to palette mode
In-Reply-To: <d7f793bd0903131503r4378cddfhe1cd9eac517a0bd0@mail.gmail.com>
References: <d7f793bd0903131503r4378cddfhe1cd9eac517a0bd0@mail.gmail.com>
Message-ID: <368a5cd50903131604n323fad24qb643c0fe6b39aafb@mail.gmail.com>

It's the same fix as this one, right?

http://bitbucket.org/effbot/pil-2009-raclette/changeset/577ae0c2a85e/

I haven't been able to force the leak during testing (the loop can
terminate in several ways, and only one of them leaks memory; none of
my test images seemed to trigger this behaviour with the test program
I used), but the fix is "obviously correct" so it's in.

</F>

From darkgl0w at yahoo.com  Fri Mar 13 22:39:36 2009
From: darkgl0w at yahoo.com (Cristi Constantin)
Date: Fri, 13 Mar 2009 14:39:36 -0700 (PDT)
Subject: [Image-SIG] Python 3 support
Message-ID: <802770.33034.qm@web52107.mail.re2.yahoo.com>

Good day.

I just joined the mailing list.

I want to know if PIL will ever support Python 3. I am slowly migrating my applications on Python 3 and i really need PIL.
It's not good to have half of my apps in Py 2.x and the rest in Py 3.x :)

Thank you in advance.


      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090313/4bc4b49d/attachment.htm>

From fredrik at pythonware.com  Sat Mar 14 00:22:02 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 14 Mar 2009 00:22:02 +0100
Subject: [Image-SIG] [raclette] 1.1.7 status
Message-ID: <368a5cd50903131622s698b57duafbca79ba5483863@mail.gmail.com>

> It's the same fix as this one, right?

By the way, here's a rough version history for the current 1.1.7
snapshot available from here:

    http://bitbucket.org/effbot/pil-2009-raclette/src/874f7005c34c/

Let me know if I've missed some patch that absolutely should go into
1.1.7, or some bug that absolutely needs to be fixed.

cheers /F

=== changes in PIL 1.1.7 as of 2009-03-13 ===

Added version attributes to jpeg/zlib/littlecms/freetype bindings, for
debugging.

Added (experimental) ImageCms function for fetching the ICC profile
for the current display (currently Windows only).

  Added HWND/HDC support to ImageCms.get_display_profile().

Added WMF renderer (Windows only).

Added ImagePointHandler and ImageTransformHandler mixins; made
ImageCmsTransform work with im.point.

Fixed potential endless loop in the XVThumbnail reader (from Nikolai
Ugelvik).

Added Kevin Cazabon's pyCMS package.

  The C code has been moved to _imagingcms.c, the Python interface
  module is installed as PIL.ImageCMS.

  Added support for in-memory ICC profiles.

  Unified buildTransform and buildTransformFromOpenProfiles.

  The profile can now be either a filename, a profile object, or a
  file-like object containing an in-memory profile.

  Additional fixes from Florian B?ch:

    Very nice - it just needs LCMS flags support so we can use black
    point compensation and softproofing :) See attached patches.  They
    also fix a naming issue which could cause confusion - display
    profile (ImageCms wording) actually means proof profile (lcms
    wording), so I changed variable names and docstrings where
    applicable. Patches are tested under Python 2.6.

Improved support for layer names in PSD files (from Sylvain Baubeau)

  Sylvain writes: I needed to be able to retrieve the names of the
  layers in a PSD files. But PsdImagePlugin.py didn't do the job so I
  wrote this very small patch.

Improved RGBA support for ImageTk for 8.4 and newer (from Con Radchenko).

  This replaces the slow run-length based encoding model with true
  compositing at the Tk level.

Added support for 16- and 32-bit images to McIdas loader.

  Based on file samples and stand-alone reader code provided by Craig
  Swank.

Added ImagePalette support to putpalette.

Fixed problem with incremental parsing of PNG files.

Make selftest.py report non-zero status on failure (from Mark Sienkiewicz)

Add big endian save support and multipage infrastructure to the TIFF
writer (from Sebastian Haase).

Added zTXT support (from Andrew Kuchling via Lowell Alleman).

Fixed potential infinite loop bug in ImageFont (from Guilherme Polo).

Added sample ICC profiles (from Kevin Cazabon)

Fixed array interface for I, F, and RGBA/RGBX images.

Added Chroma subsampling support for JPEG (from Justin Huff).

  Justin writes: Attached is a patch (against PIL 1.1.6) to provide
  control over the chroma subsampling done by the JPEG encoder.  This
  is often useful for reducing compression artifacts around edges of
  clipart and text.

Added USM/Gaussian Blur code from Kevin Cazabon.

Fixed bug w. uninitialized image data when cropping outside the source image.

Use ImageShow to implement the Image.show method.

  Most notably, this picks the 'display' utility when available.  It
  also allows application code to register new display utilities via
  the ImageShow registry.

Release the GIL in the PNG compressor (from Michael van Tellingen).

Revised JPEG CMYK handling.

  Always assume Adobe behaviour, both when reading and writing (based on
  a patch by Kevin Cazabon, and test data by Tim V. and Charlie Clark).

Support for preserving ICC profiles (by Florian B?ch via Tim Hatch).

  Florian writes:

  It's a beta, so still needs some testing, but should allow you to:
  - retain embedded ICC profiles when saving from/to JPEG, PNG, TIFF.
     Existing code doesn't need to be changed.
  - access embedded profiles in JPEG, PNG, PSD, TIFF.

  It also includes patches for TIFF to retain IPTC, Photoshop and XMP
  metadata when saving as TIFF again, read/write TIFF resolution
  information correctly, and to correct inverted CMYK JPEG files.

Fixed potential memory leak in median cut quantizer (from Evgeny Salmin).

Fixed OverflowError when reading upside-down BMP images.

Added resolution save option for PDF files.

  Andreas Kostyrka writes: I've included a patched PdfImagePlugin.py
  based on 1.1.6 as included in Ubuntu, that supports a "resolution"
  save option. Not great, but it makes the PDF saving more useful by
  allowing PDFs that are not exactly 72dpi.

Look for Tcl/Tk include files in version-specific include directory
(from Encolpe Degoute).

Fixed grayscale rounding error in ImageColor.getcolor (from Tim
Hatch).

Fixed truetype positioning when first character has a negative left
bearing (from Ned Batchelder):

  Ned writes: In PIL 1.1.6, ImageDraw.text will position the string
  incorrectly if the first character has a negative left bearing.  To
  see the problem, show a string like "///" in an italic font.  The
  first slash will be clipped at the left, and the string will be
  mis-positioned.

Fixed resolution unit bug in tiff reader/writer (from Florian H?ch,
Gary Bloom, and others).

Added simple transparency support for RGB images (reported by
Sebastian Spaeth).

Added support for Unicode filenames in ImageFont.truetype (from Donn
Ingle).

Fixed potential crash in ImageFont.getname method (from Donn Ingle).

Fixed encoding issue in PIL/WalImageFile (from Santiago M. Mola).

=== end ===

From cannon.el at gmail.com  Sat Mar 14 01:49:05 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Fri, 13 Mar 2009 17:49:05 -0700
Subject: [Image-SIG] perspective method
Message-ID: <b986b290903131749h246f165bk78f9b5d35895d994@mail.gmail.com>

A while back a read about, and even used, an experimental feature in
PIL 1.1.16 it was a new feature to the im.transform() method. as I
recall it was im.transform(size, PERSPECTIVE, data, filter) where data
was the transformation matrix. I was wondering what the status of this
feature was. It worked fine for me, and I wondered if it was still
considered experimental or if we could now consider it stable and add
it to the documentation.
Edward

From lists+Image_SIG at hoech.org  Sat Mar 14 09:53:16 2009
From: lists+Image_SIG at hoech.org (=?ISO-8859-1?Q?Florian_H=F6ch?=)
Date: Sat, 14 Mar 2009 09:53:16 +0100
Subject: [Image-SIG] [raclette] 1.1.7 status
In-Reply-To: <368a5cd50903131622s698b57duafbca79ba5483863@mail.gmail.com>
References: <368a5cd50903131622s698b57duafbca79ba5483863@mail.gmail.com>
Message-ID: <49BB707C.3030506@hoech.org>

I wrote a patch for CMYK undercolor removal for the "dumb" (non-ICC) 
conversion a while ago, but I think I never posted it anywhere - so here 
goes (it works flawlessly as far as I can tell, but I as I'm no seasoned 
C-programmer the code is maybe not very efficient). Only drawback I can 
think of is that the results are of course inconsistent with previous 
PIL versions, but you gain the benefit that the conversion can be easily 
inverted.

Something else I'd like to see is 16-bits-per-channel support for image 
types that can do it (hmm, TIFF, PSD?), but I have no idea how much work 
that'd be, so maybe not for 1.1.7.

Regards,

Florian H?ch

Fredrik Lundh wrote:
> By the way, here's a rough version history for the current 1.1.7
> snapshot available from here:
> 
>     http://bitbucket.org/effbot/pil-2009-raclette/src/874f7005c34c/
> 
> Let me know if I've missed some patch that absolutely should go into
> 1.1.7, or some bug that absolutely needs to be fixed.
> 
> cheers /F
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Convert.c.diff
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090314/a9cdc976/attachment.txt>

From v+python at g.nevcal.com  Sat Mar 14 17:20:25 2009
From: v+python at g.nevcal.com (Glenn Linderman)
Date: Sat, 14 Mar 2009 09:20:25 -0700
Subject: [Image-SIG] [raclette] 1.1.7 status
In-Reply-To: <49BB707C.3030506@hoech.org>
References: <368a5cd50903131622s698b57duafbca79ba5483863@mail.gmail.com>
	<49BB707C.3030506@hoech.org>
Message-ID: <49BBD949.7050506@g.nevcal.com>

On approximately 3/14/2009 1:53 AM, came the following characters from 
the keyboard of Florian H?ch:
> Something else I'd like to see is 16-bits-per-channel support for 
> image types that can do it (hmm, TIFF, PSD?), but I have no idea how 
> much work that'd be, so maybe not for 1.1.7.

And I'd like 1-bit-per-1-channel support (bitonal) for TIFF and Group4 
compression on writing.... guess we do different stuff, eh?

-- 
Glenn -- http://nevcal.com/
===========================
A protocol is complete when there is nothing left to remove.
-- Stuart Cheshire, Apple Computer, regarding Zero Configuration Networking


From lists+Image_SIG at hoech.org  Sat Mar 14 17:28:33 2009
From: lists+Image_SIG at hoech.org (=?ISO-8859-1?Q?Florian_H=F6ch?=)
Date: Sat, 14 Mar 2009 17:28:33 +0100
Subject: [Image-SIG] [raclette] 1.1.7 status
In-Reply-To: <49BBD949.7050506@g.nevcal.com>
References: <368a5cd50903131622s698b57duafbca79ba5483863@mail.gmail.com>
	<49BB707C.3030506@hoech.org> <49BBD949.7050506@g.nevcal.com>
Message-ID: <49BBDB31.9060504@hoech.org>

Maybe you shouldn't have mentioned it - now I'd like to have that too ;)

Regards,

Florian H?ch

Glenn Linderman wrote:
> And I'd like 1-bit-per-1-channel support (bitonal) for TIFF and Group4 
> compression on writing.... guess we do different stuff, eh?


From evg.salmin at gmail.com  Sat Mar 14 11:10:29 2009
From: evg.salmin at gmail.com (Evgeny Salmin)
Date: Sat, 14 Mar 2009 16:10:29 +0600
Subject: [Image-SIG] Converting images to palette mode
In-Reply-To: <368a5cd50903131604n323fad24qb643c0fe6b39aafb@mail.gmail.com>
References: <d7f793bd0903131503r4378cddfhe1cd9eac517a0bd0@mail.gmail.com>
	<368a5cd50903131604n323fad24qb643c0fe6b39aafb@mail.gmail.com>
Message-ID: <d7f793bd0903140310of77efav4ec00bb7f096364b@mail.gmail.com>

Hello Fredrik.

Yes, it's the same fix. I just did't know that you've added fix from my
previous post into a repository. Is it official PIL repo?

2009/3/14 Fredrik Lundh <fredrik at pythonware.com>

> It's the same fix as this one, right?
>
> http://bitbucket.org/effbot/pil-2009-raclette/changeset/577ae0c2a85e/
>
> I haven't been able to force the leak during testing (the loop can
> terminate in several ways, and only one of them leaks memory; none of
> my test images seemed to trigger this behaviour with the test program
> I used), but the fix is "obviously correct" so it's in.
>
> </F>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090314/143140ba/attachment.htm>

From mark.wendell at gmail.com  Sun Mar 15 18:15:08 2009
From: mark.wendell at gmail.com (Mark Wendell)
Date: Sun, 15 Mar 2009 11:15:08 -0600
Subject: [Image-SIG] ImageMath power-function error
Message-ID: <b22700f10903151015s737c7523oc720aa9c46420228@mail.gmail.com>

Hey all,

I can't seem to use the power ("**") operator in the ImageMath module.


For example, multiplying an L-mode image by a factor of 2.0 works fine:

>>> out = ImageMath.eval( "convert( ( a*val ), 'L')", a=im, val=2.0)
(no errors)

However, raising the same image by a power of 2.0 generates a bad operand
type error:

>>> out = ImageMath.eval( "convert( ( a**val ), 'L')", a=im, val=2.0)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python2.5/site-packages/PIL/ImageMath.py", line 203, in
eval
    out =__builtin__.eval(expression, args)
  File "<string>", line 1, in <module>
  File "/usr/lib/python2.5/site-packages/PIL/ImageMath.py", line 119, in
__pow__
    return self.apply("pow", self, other)
  File "/usr/lib/python2.5/site-packages/PIL/ImageMath.py", line 82, in
apply
    raise TypeError, "bad operand type for '%s'" % op
TypeError: bad operand type for 'pow'


Am I making a silly mistake?

thanks in advance,
Mark

Ubuntu 8.10 linux, python 2.5.2, Pil 1.1.6
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090315/82e78587/attachment.htm>

From pander at users.sourceforge.net  Mon Mar 16 11:21:56 2009
From: pander at users.sourceforge.net (Pander)
Date: Mon, 16 Mar 2009 11:21:56 +0100
Subject: [Image-SIG] Version 0.3 of ImageText.py
Message-ID: <49BE2844.4080708@users.sourceforge.net>

Hi all,

Here is version 0.3 of ImageText.py. I hope it can find a way into the
releases of PIL or contribute to PIL uses one wa or the other. This
version has lots of improvements suggested by PyLint and Python 2.6.
Have fun with it.

Regards,

Pander
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ImageText.py
Type: text/x-python
Size: 15551 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090316/8400cdd7/attachment.py>

From fredrik at pythonware.com  Mon Mar 16 15:01:48 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 16 Mar 2009 15:01:48 +0100
Subject: [Image-SIG] Converting images to palette mode
In-Reply-To: <d7f793bd0903140310of77efav4ec00bb7f096364b@mail.gmail.com>
References: <d7f793bd0903131503r4378cddfhe1cd9eac517a0bd0@mail.gmail.com>
	<368a5cd50903131604n323fad24qb643c0fe6b39aafb@mail.gmail.com>
	<d7f793bd0903140310of77efav4ec00bb7f096364b@mail.gmail.com>
Message-ID: <368a5cd50903160701r51a9b589i5715f7662a91cd97@mail.gmail.com>

Official for the 1.1.7 release, at least.  Not sure what we'll do
beyond that (and whatever we do, the finished release will be
available from svn.effbot.org for svn:externals users, as usual).

</F>

On Sat, Mar 14, 2009 at 11:10 AM, Evgeny Salmin <evg.salmin at gmail.com> wrote:
> Hello Fredrik.
>
> Yes, it's the same fix. I just did't know that you've added fix from my
> previous post into a repository. Is it official PIL repo?

From fredrik at pythonware.com  Mon Mar 16 15:05:33 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 16 Mar 2009 15:05:33 +0100
Subject: [Image-SIG] ImageMath power-function error
In-Reply-To: <b22700f10903151015s737c7523oc720aa9c46420228@mail.gmail.com>
References: <b22700f10903151015s737c7523oc720aa9c46420228@mail.gmail.com>
Message-ID: <368a5cd50903160705u6477a226vba79e40f0e6cf420@mail.gmail.com>

No time to double-check, but I suspect the 1.1.6 version doesn't
support ** - it's mentioned in the documentation, but I don't see any
sign of it in the source code.  It's on my list of things to fix for
1.1.7.

</F>

2009/3/15 Mark Wendell <mark.wendell at gmail.com>:
> Hey all,
>
> I can't seem to use the power ("**") operator in the ImageMath module.
>
>
> For example, multiplying an L-mode image by a factor of 2.0 works fine:
>
>>>> out = ImageMath.eval( "convert( ( a*val ), 'L')", a=im, val=2.0)
> (no errors)
>
> However, raising the same image by a power of 2.0 generates a bad operand
> type error:
>
>>>> out = ImageMath.eval( "convert( ( a**val ), 'L')", a=im, val=2.0)
> Traceback (most recent call last):
> ? File "<stdin>", line 1, in <module>
> ? File "/usr/lib/python2.5/site-
> packages/PIL/ImageMath.py", line 203, in eval
> ??? out =__builtin__.eval(expression, args)
> ? File "<string>", line 1, in <module>
> ? File "/usr/lib/python2.5/site-packages/PIL/ImageMath.py", line 119, in
> __pow__
> ??? return self.apply("pow", self, other)
> ? File "/usr/lib/python2.5/site-packages/PIL/ImageMath.py", line 82, in
> apply
> ??? raise TypeError, "bad operand type for '%s'" % op
> TypeError: bad operand type for 'pow'
>
>
> Am I making a silly mistake?
>
> thanks in advance,
> Mark
>
> Ubuntu 8.10 linux, python 2.5.2, Pil 1.1.6
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>

From cfp at centeris2009.eiswatch.org  Mon Mar 16 14:01:58 2009
From: cfp at centeris2009.eiswatch.org (CENTERIS'2009 - Conference on ENTERprise Information Systems)
Date: Mon, 16 Mar 2009 13:01:58 +0000
Subject: [Image-SIG] 2nd Call for Papers | Extended deadline: April 07 |
 CENTERIS'2009 - Conference on ENTERprise Information Systems | 2nd Call for
 Papers
Message-ID: <20090316141853.D98101E4002@bag.python.org>

-------------
Please consider contributing to and/or forwarding to the appropriate groups and peers the following call for papers.
(please excuse us if you received this call more than once)
-------------
You are receiving this email because of your research activities on the conference topic. 
To unsubscribe please send an email to secretariat at centeris.eiswatch.org with the subject "Unsubscribe image-sig at python.org" 
-------------


------------- CENTERIS?2009  |  Call for Papers 
------------- Conference on ENTERprise Information Systems

------------- Ofir, Portugal, 07-09 October 2009

http://centeris.eiswatch.org
extended submission deadline (full paper):  April 07, 2009
-------------


Dear Sirs,

It is our great pleasure to invite you to the CENTERIS?2009 - Conference on ENTERprise Information Systems ? aligning technology, organizations and people, to be held in Ofir, Portugal, organized by the Polytechnic Institute of C?vado and Ave and the University of Tr?s-os-Montes e Alto Douro.

This is the place where, from 07 to 09 October 2009, under the leitmotiv of Enterprise Information Systems, academics, scientists, IT/IS professionals, managers and solution providers from all over the world will have the opportunity to share experiences, bring new ideas, debate issues, introduce the latest developments in the field.

CENTERIS?2009 is an international conference that will address the largely multidisciplinary field embraced by the Enterprise Information Systems (EIS), from the social, organizational and technological perspectives. 
The 3-day conference follows an innovative format, previewing scientific sessions, tutorial and technical sessions, and simultaneously exhibitions by solutions providers and product presentations. 
There will also be a social program that will enable participants to get to know each other and informally exchange information, contacts and experiences, discuss projects, and incubate partnerships.

-------------
------------- Submission and Publications

Submissions will be reviewed on a double-blind review basis. 
Only original contributions will be accepted.

All accepted papers will be published in the Conference Proceedings Book (full texts) with ISBN and on a CD-ROM to be distributed during the conference.

Authors of a selection of around 25 papers will be invited for enhancing and extending their papers for inclusion in a book to be published by IGI-Global, to be released by early 2010. Depending on the number of high quality papers accepted, the Organization can consider the edition of two books, instead of one.

According to the Scientific Committee recommendations, authors of selected papers will be invited to enhance the paper for publication in:

     -  A Special Issue of the Information Resources Management Journal (IRMJ);
     -  The International Journal of Enterprise Information Systems (IJEIS);
     -  Journal of Theoretical and Applied Electronic Commerce Research (JTAER);
     -  International Journal of Information Technologies and Systems Approach (IJITSA);
     -  The emerging ?Information and Communication Technologies for the Advanced Enterprise: an international journal? (ICT'ae).

At the end of this message you will find the composition of the Editorial Board.

------------- 
------------- Hotel and travel arrangements

The Conference will take place at AXIS OFIR Beach Resort Hotel in Ofir, Esposende, Portugal, a 4-star Hotel located in one of the most beautiful landscapes in the North of Portugal. The Atlantic coast of Esposende is part of the Natural Park of the North Coast, a protected area of great beauty.
Ofir is located in a wonderful historic and cultural region with three locations/zones proclaimed as UNESCO cultural World Heritage within 100 km: Historic Centre of Guimaraes (50Km), Historic Centre of Porto (40Km) and Alto Douro Wine Region.
The Hotel Axis Ofir is located 35 km north of the Oporto International Airport ?Francisco Sa Carneiro?. It is connected by the highway A28. We suggest getting a taxi which cost would be around 35 Euros (one way), taking less than 30 min to the Hotel.
------------- 

For more detailed information, please visit  http://centeris.eiswatch.org

We hope to see you in Ofir.

Best wishes,
Manuela Cunha
Lu?s Amaral
Jo?o Varaj?o


------------- General Chair
Maria Manuela Cruz-Cunha  (mcunha at ipca.pt)
Polytechnic Institute of C?vado and Ave, Portugal

------------- Program Chair
Lu?s Amaral  (amaral at dsi.uminho.pt)
University of Minho, Portugal

------------- Organization Chair
Jo?o Eduardo Varaj?o  (jvarajao at utad.pt)
University of Tr?s-os-Montes e Alto Douro, Portugal

------------- Secretariat
secretariat at centeris.eiswatch.org



------------- 
------------- Editorial Board
(This list is not complete as the invitation process is still going on)

Adamantios Koumpis, Research Programmes Division, ALTEC S.A, Grece
Adolfo Vanti, Universidade do Vale do Rio dos Sinos, Brasil
Albert Boonstra, University of Groningen, The Netherlands
Alberto Arroyo, Orienta, Spain
Alexandra Klen, Universidade Federal de Santa Catarina, Brazil
?lvaro Rocha, University Fernando Pessoa, Portugal
Ana Maria Fermoso Garcia, Pontifical University of Salamanca, Spain
Anca Draghici, Politehnica University of Timisoara, Romania
Andr?a Paiva, Universidade de S?o Paulo, Brazil 
Angappa Gunasekaran, University of Massachusetts Dartmouth, USA
Antonio Guevara, University of Malaga, Spain
Antonio Jos? Balloni, CTI, Brasil
Ant?nio Trigo, Polytechnic Institute of Coimbra, Portugal
Aysin Rahimifard, Loughborough University, UK
Bart H.M. Gerritsen, TNO N. Org. for App. Scientific Research, The Netherlands
Calin Gurau, GSCM ? Montpellier Business School, France
Carlos Ferr?s Sexto, Universidad de Santiago de Compostela, Spain
Carlos Machado-Santos, University of Tr?s-os-Montes e Alto, Portugal
Carmen de Pablos, Rey Juan Carlos University, Spain
Carola Jones, Universidad Nacional de C?rdoba, Argentina
Carrillo Verd?n, Universidad Polit?cnica de Madrid, Spain
Cesar Alexandre de Souza, University of Sao Paulo, Brazil
Chad Lin, Curtin University of Technology, Australia
Daniel L?bke, Leibniz University Hannover, Germany
Dimitrios Koufopoulos, Brunel University, UK
Dirk Schaefer, Georgia Institute of Technology Savannah, USA
Dirk Werth, Institut f?r Wirtschaftsinformatik, Germany
Dulce Domingos, University of Lisbon, Portugal
Duminda Wijesekera, George Mason University, USA
Edson Luiz Riccio, University of Sao Paulo, Brasil
Efrem Mallach, University of Massachusetts Dartmouth, USA
Eitel Lauria, Marist College, USA
Enrique Paniagua Ar?s, Universidad de Murcia, Spain
Ercan Oztemel, Marmara University, Turkey
Esra Kurt Tekez, Sakarya University, Turkey
Fahd Alizai, Victoria University Australia, Australia
Filomena Lopes, Universidade Portucalense, Portugal
Gabor Hosszu, Budpest University of Technology and Economics, Hungary
George Ioannou, Athens Faculty of Economics and Business, Greece
George Leal Jamil, FUMEC/BH, Brasil
Giorgio Bruno, Politecnico di Torino, Italy 
Goran Putnik, University of Minho, Portugal
Heiko Duin, BIBA Bremer Institut f?r Produktion und Logistik GmbH, Germany
Henrique O?Neill, ISCTE, Portugal
Igor Perko, University of Maribor, Slovenia
llan Oshri, Rotterdam School of Management, The Netherlands
Isabel Ramos, University of Minho, Portugal
Ivanilde Eyng, UEPG/UNAM, Brasil
Jaideep Motwani, Grand Valley State University, USA
Jaime Munoz, Autonomous University of Aguascalientes, Mexico
Jens Eschenb?cher, BIBA Bremer Institut f?r Produktion und Logistik, Germany
Jerzy Kisielnicki, University of Warsaw, Poland
Joao Barroso, University of Tr?s-os-Montes e Alto Douro, Portugal
Joao Varajao, University of Tr?s-os-Montes e Alto Douro, Portugal
Jorge Marx G?mez, Oldenburg University, Germany
Jos? Bulas Cruz, University of Tr?s-os-Montes e Alto Douro, Portugal
Jos? Carlos Metr?lho, Polytechnic Institute of Castelo Branco, Portugal
Jos? L. Caro, University of Malaga, Spain
Jos? L. Leiva, University of Malaga, Spain
Jos? Luis Mota Pereira, University of Minho, Portugal
Kam Hou Vat, University of Macau, Macau
Klara Antlova, Technical university of Liberec, Czech republic
Leonardo Soto, University of Guadalajara, Mexico
Leonel Santos, University of Minho, Portugal
Lorna Uden, Staffordshire University, UK
Lu?s Amaral, University of Minho, Portugal
Lu?s Borges Gouveia, University Fernando Pessoa, Portugal
Mahesh S. Raisinghani, Texas Woman's University, USA
Malihe Tabatabaie, University of York, UK
Manuel Filipe Santos, University of Minho, Portugal
Manuel Jo?o Pereira, Universidade Cat?lica, Portugal
Manuel Mora, Autonomous University of Aguascalientes, Mexico
Manuel P?rez Cota, Universidad de Vigo, Spain
Maria Argyropoulou, Brunel University, UK
Maria Manuela Cruz-Cunha, Polytechnic Institute of C?vado and Ave, Portugal
Marco Kuhrmann, Technische Universit?t M?nchen, Germany
Marilisa Oliveira, UEPG/UNAM, Brazil
Marko Kolakovic, University of Zagreb, Croatia
M?rio Caldeira, Technical University of Lisbon, Portugal
Mayumi Hori, Hakuoh University, Japan
Michal ?emlicka, Charles University, Czech Republic
Mirjana Stojanovic, University of Belgrade, Serbia
Narciso Cerpa, University of Talca, Chile
Ovsei Gelman, CCADET-UNAM, Mexico
?zalp Vayvay, Faculty of Engineering Marmara University, Turkey
Ozden Ustun, Dumlupinar University, Turkey
Patr?cia Gon?alves, Polytechnic Institute of C?vado and Ave, Portugal
Paulo Martins, University of Tr?s-os-Montes e Alto Douro, Portugal
Paula Morais, Universidade Portucalense, Portugal
Paulo Rupino, University of Coimbra, Portugal
Paulo Tom?, Polytechnic Institute of Viseu, Portugal
Pedro Ara?jo, University of Beira Interior, Portugal
Pedro Campos, University of Porto, Portugal
Pedro Quelhas de Brito, University of Porto, Portugal
Pedro Ribeiro, University of Minho, Portugal
Pedro Soto Acosta, Universidad de Murcia, Spain
Polo, University of Sao Paulo, Brasil
Protogeros Nicolaos, University of Macedonia Economic and Social Scs., Greece
Ramiro Gon?alves, University of Tr?s-os-Montes e Alto Douro, Portugal
Ricardo Colomo Palacios, University Carlos III of Madrid, Spain
Ricardo Gon?alves, Universidade Nova de Lisboa, Portugal
Ricardo Simoes, Polytechnic Institute of C?vado and Ave, Portugal
Richard Burkhard, San Jose State University, USA
Rinaldo Michelini, PMAR Lab of the University of  Genova, Italy
Roberto Razzoli, PMAR Lab of the University of Genova, Italy
Rui Dinis Sousa, University of Minho, Portugal
Rui Pedro Rijo, Polytechnic Institute of Leiria, Portugal
Samo Bobek, University of Maribor, Slovenia
Sanja Vranes, The Mihajlo Pupin Institute, Serbia
Simona Sternad, University of Maribor, Slovenia
Snezana Pantelic, The Mihajlo Pupin Institute, Serbia
Subhasish Dasgupta, George Washington University, USA
Tahinakis Panayiotis, University of Macedonia Economic and Social Scs., Greece
Valentina Janev, The Mihajlo Pupin Institute, Serbia
Vladanka Acimovic-Raspopovic, University of Belgrade, Serbia
Vitor Basto Fernandes, Polytechnic Institute of Leiria, Portugal
V?tor Carvalho, Polytechnic Institute of C?vado and Ave, Portugal
V?tor Santos, Microsoft, Portugal
Vladim?r Modr?k, Technical University of Kosice, Slovakia
Vojko Potocan, University of Maribor, Slovenia
Wai Ming Cheung, University of Bath, UK


From mark.wendell at gmail.com  Mon Mar 16 15:09:48 2009
From: mark.wendell at gmail.com (Mark Wendell)
Date: Mon, 16 Mar 2009 08:09:48 -0600
Subject: [Image-SIG] ImageMath power-function error
In-Reply-To: <368a5cd50903160705u6477a226vba79e40f0e6cf420@mail.gmail.com>
References: <b22700f10903151015s737c7523oc720aa9c46420228@mail.gmail.com>
	<368a5cd50903160705u6477a226vba79e40f0e6cf420@mail.gmail.com>
Message-ID: <b22700f10903160709m37261bbqfb0a6c39df477f73@mail.gmail.com>

Hmm. The 1.1.6 docs for the ImageMath module say "The module also supports
unary minus (-), modulo (%), and power (**) operators."

Still, if not, looking forward to having the functionality in 1.1.7.

thanks
Mark

On Mon, Mar 16, 2009 at 8:05 AM, Fredrik Lundh <fredrik at pythonware.com>wrote:

> No time to double-check, but I suspect the 1.1.6 version doesn't
> support ** - it's mentioned in the documentation, but I don't see any
> sign of it in the source code.  It's on my list of things to fix for
> 1.1.7.
>
> </F>
>
> 2009/3/15 Mark Wendell <mark.wendell at gmail.com>:
> > Hey all,
> >
> > I can't seem to use the power ("**") operator in the ImageMath module.
> >
> >
> > For example, multiplying an L-mode image by a factor of 2.0 works fine:
> >
> >>>> out = ImageMath.eval( "convert( ( a*val ), 'L')", a=im, val=2.0)
> > (no errors)
> >
> > However, raising the same image by a power of 2.0 generates a bad operand
> > type error:
> >
> >>>> out = ImageMath.eval( "convert( ( a**val ), 'L')", a=im, val=2.0)
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in <module>
> >   File "/usr/lib/python2.5/site-
> > packages/PIL/ImageMath.py", line 203, in eval
> >     out =__builtin__.eval(expression, args)
> >   File "<string>", line 1, in <module>
> >   File "/usr/lib/python2.5/site-packages/PIL/ImageMath.py", line 119, in
> > __pow__
> >     return self.apply("pow", self, other)
> >   File "/usr/lib/python2.5/site-packages/PIL/ImageMath.py", line 82, in
> > apply
> >     raise TypeError, "bad operand type for '%s'" % op
> > TypeError: bad operand type for 'pow'
> >
> >
> > Am I making a silly mistake?
> >
> > thanks in advance,
> > Mark
> >
> > Ubuntu 8.10 linux, python 2.5.2, Pil 1.1.6
> > _______________________________________________
> > Image-SIG maillist  -  Image-SIG at python.org
> > http://mail.python.org/mailman/listinfo/image-sig
> >
> >
>



-- 
--
Mark Wendell
mark.wendell at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090316/e7b1990e/attachment-0001.htm>

From irgendetwas at charlie.at  Tue Mar 17 09:00:03 2009
From: irgendetwas at charlie.at (charlie)
Date: Tue, 17 Mar 2009 09:00:03 +0100
Subject: [Image-SIG] [raclette] 1.1.7 status
In-Reply-To: <368a5cd50903131622s698b57duafbca79ba5483863@mail.gmail.com>
References: <368a5cd50903131622s698b57duafbca79ba5483863@mail.gmail.com>
Message-ID: <37554e120903170100r211c0090qe4a05734d9a5a2f3@mail.gmail.com>

Hi Fredrik,

would you consider Eric's patch:
http://mail.python.org/pipermail/image-sig/2007-March/004359.html
for supporting multiple instances of a tag in the iptc info (ie 2:25 keywords)

thanks in advance
  charlie

On Sat, Mar 14, 2009 at 12:22 AM, Fredrik Lundh <fredrik at pythonware.com> wrote:
>> It's the same fix as this one, right?
>
> By the way, here's a rough version history for the current 1.1.7
> snapshot available from here:
>
> ? ?http://bitbucket.org/effbot/pil-2009-raclette/src/874f7005c34c/
>
> Let me know if I've missed some patch that absolutely should go into
> 1.1.7, or some bug that absolutely needs to be fixed.
>
> cheers /F
>
> === changes in PIL 1.1.7 as of 2009-03-13 ===
>
> Added version attributes to jpeg/zlib/littlecms/freetype bindings, for
> debugging.
>
> Added (experimental) ImageCms function for fetching the ICC profile
> for the current display (currently Windows only).
>
> ?Added HWND/HDC support to ImageCms.get_display_profile().
>
> Added WMF renderer (Windows only).
>
> Added ImagePointHandler and ImageTransformHandler mixins; made
> ImageCmsTransform work with im.point.
>
> Fixed potential endless loop in the XVThumbnail reader (from Nikolai
> Ugelvik).
>
> Added Kevin Cazabon's pyCMS package.
>
> ?The C code has been moved to _imagingcms.c, the Python interface
> ?module is installed as PIL.ImageCMS.
>
> ?Added support for in-memory ICC profiles.
>
> ?Unified buildTransform and buildTransformFromOpenProfiles.
>
> ?The profile can now be either a filename, a profile object, or a
> ?file-like object containing an in-memory profile.
>
> ?Additional fixes from Florian B?ch:
>
> ? ?Very nice - it just needs LCMS flags support so we can use black
> ? ?point compensation and softproofing :) See attached patches. ?They
> ? ?also fix a naming issue which could cause confusion - display
> ? ?profile (ImageCms wording) actually means proof profile (lcms
> ? ?wording), so I changed variable names and docstrings where
> ? ?applicable. Patches are tested under Python 2.6.
>
> Improved support for layer names in PSD files (from Sylvain Baubeau)
>
> ?Sylvain writes: I needed to be able to retrieve the names of the
> ?layers in a PSD files. But PsdImagePlugin.py didn't do the job so I
> ?wrote this very small patch.
>
> Improved RGBA support for ImageTk for 8.4 and newer (from Con Radchenko).
>
> ?This replaces the slow run-length based encoding model with true
> ?compositing at the Tk level.
>
> Added support for 16- and 32-bit images to McIdas loader.
>
> ?Based on file samples and stand-alone reader code provided by Craig
> ?Swank.
>
> Added ImagePalette support to putpalette.
>
> Fixed problem with incremental parsing of PNG files.
>
> Make selftest.py report non-zero status on failure (from Mark Sienkiewicz)
>
> Add big endian save support and multipage infrastructure to the TIFF
> writer (from Sebastian Haase).
>
> Added zTXT support (from Andrew Kuchling via Lowell Alleman).
>
> Fixed potential infinite loop bug in ImageFont (from Guilherme Polo).
>
> Added sample ICC profiles (from Kevin Cazabon)
>
> Fixed array interface for I, F, and RGBA/RGBX images.
>
> Added Chroma subsampling support for JPEG (from Justin Huff).
>
> ?Justin writes: Attached is a patch (against PIL 1.1.6) to provide
> ?control over the chroma subsampling done by the JPEG encoder. ?This
> ?is often useful for reducing compression artifacts around edges of
> ?clipart and text.
>
> Added USM/Gaussian Blur code from Kevin Cazabon.
>
> Fixed bug w. uninitialized image data when cropping outside the source image.
>
> Use ImageShow to implement the Image.show method.
>
> ?Most notably, this picks the 'display' utility when available. ?It
> ?also allows application code to register new display utilities via
> ?the ImageShow registry.
>
> Release the GIL in the PNG compressor (from Michael van Tellingen).
>
> Revised JPEG CMYK handling.
>
> ?Always assume Adobe behaviour, both when reading and writing (based on
> ?a patch by Kevin Cazabon, and test data by Tim V. and Charlie Clark).
>
> Support for preserving ICC profiles (by Florian B?ch via Tim Hatch).
>
> ?Florian writes:
>
> ?It's a beta, so still needs some testing, but should allow you to:
> ?- retain embedded ICC profiles when saving from/to JPEG, PNG, TIFF.
> ? ? Existing code doesn't need to be changed.
> ?- access embedded profiles in JPEG, PNG, PSD, TIFF.
>
> ?It also includes patches for TIFF to retain IPTC, Photoshop and XMP
> ?metadata when saving as TIFF again, read/write TIFF resolution
> ?information correctly, and to correct inverted CMYK JPEG files.
>
> Fixed potential memory leak in median cut quantizer (from Evgeny Salmin).
>
> Fixed OverflowError when reading upside-down BMP images.
>
> Added resolution save option for PDF files.
>
> ?Andreas Kostyrka writes: I've included a patched PdfImagePlugin.py
> ?based on 1.1.6 as included in Ubuntu, that supports a "resolution"
> ?save option. Not great, but it makes the PDF saving more useful by
> ?allowing PDFs that are not exactly 72dpi.
>
> Look for Tcl/Tk include files in version-specific include directory
> (from Encolpe Degoute).
>
> Fixed grayscale rounding error in ImageColor.getcolor (from Tim
> Hatch).
>
> Fixed truetype positioning when first character has a negative left
> bearing (from Ned Batchelder):
>
> ?Ned writes: In PIL 1.1.6, ImageDraw.text will position the string
> ?incorrectly if the first character has a negative left bearing. ?To
> ?see the problem, show a string like "///" in an italic font. ?The
> ?first slash will be clipped at the left, and the string will be
> ?mis-positioned.
>
> Fixed resolution unit bug in tiff reader/writer (from Florian H?ch,
> Gary Bloom, and others).
>
> Added simple transparency support for RGB images (reported by
> Sebastian Spaeth).
>
> Added support for Unicode filenames in ImageFont.truetype (from Donn
> Ingle).
>
> Fixed potential crash in ImageFont.getname method (from Donn Ingle).
>
> Fixed encoding issue in PIL/WalImageFile (from Santiago M. Mola).
>
> === end ===
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From Chris.Barker at noaa.gov  Tue Mar 17 17:28:42 2009
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Tue, 17 Mar 2009 09:28:42 -0700
Subject: [Image-SIG] ImageMath power-function error
In-Reply-To: <b22700f10903160709m37261bbqfb0a6c39df477f73@mail.gmail.com>
References: <b22700f10903151015s737c7523oc720aa9c46420228@mail.gmail.com>
	<368a5cd50903160705u6477a226vba79e40f0e6cf420@mail.gmail.com>
	<b22700f10903160709m37261bbqfb0a6c39df477f73@mail.gmail.com>
Message-ID: <49BFCFBA.1000407@noaa.gov>

Mark Wendell wrote:
> Still, if not, looking forward to having the functionality in 1.1.7.

Another option is to convert to a numpy array, and do your math there -- 
you'll get full featured math that way.

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From fredrik at pythonware.com  Tue Mar 17 22:22:01 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 17 Mar 2009 22:22:01 +0100
Subject: [Image-SIG] ANN: PIL 1.1.7 alpha 1 (March 17, 2009)
Message-ID: <368a5cd50903171422n79229f9dpf76fa19e772bbcd9@mail.gmail.com>

The first official alpha of PIL 1.1.7 is now available; you can get
source archives from

   http://bitbucket.org/effbot/pil-2009-raclette/src/pil-117-20090317

(use the download >> menu in the right corner).  Alternatively, you
can check out the alpha using Mercurial:

   $ hg clone -r pil-117-20090317 http://bitbucket.org/effbot/pil-2009-raclette/

or Subversion:

    $ svn co http://svn.effbot.org/public/tags/pil-1.1.7a1-20090317/

(note that the main HG repo contains a few odd heads and most likely a
bunch of changes done after the alpha was cut)

I hope to have (some) Windows builds available early next week, at the latest.

</F>

From ktenney at gmail.com  Wed Mar 18 14:43:02 2009
From: ktenney at gmail.com (Kent Tenney)
Date: Wed, 18 Mar 2009 08:43:02 -0500
Subject: [Image-SIG] png.save() and tEXt chunks
Message-ID: <ca816810903180643g54b6158dk60a87a46c7d7243d@mail.gmail.com>

Howdy,

I work with png's which have tEXt chunks containing metadata.
PIL discards them upon open .. save.

Might it be desirable for core save() to check for extra chunks
and preserve them if found?

Thanks,
Kent

From fetchinson at googlemail.com  Thu Mar 19 02:39:12 2009
From: fetchinson at googlemail.com (Daniel Fetchinson)
Date: Wed, 18 Mar 2009 18:39:12 -0700
Subject: [Image-SIG] Python 3 support
In-Reply-To: <802770.33034.qm@web52107.mail.re2.yahoo.com>
References: <802770.33034.qm@web52107.mail.re2.yahoo.com>
Message-ID: <fbe2e2100903181839t526548b4pd1b5fe4b8dd6b53f@mail.gmail.com>

> I just joined the mailing list.
>
> I want to know if PIL will ever support Python 3. I am slowly migrating my
> applications on Python 3 and i really need PIL.
> It's not good to have half of my apps in Py 2.x and the rest in Py 3.x :)
>
> Thank you in advance.

Please see the thread

http://mail.python.org/pipermail/python-dev/2009-March/087317.html

HTH,
Daniel

-- 
Psss, psss, put it down! - http://www.cafepress.com/putitdown

From fredrik at pythonware.com  Thu Mar 19 09:31:32 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 19 Mar 2009 09:31:32 +0100
Subject: [Image-SIG] Python 3 support
In-Reply-To: <802770.33034.qm@web52107.mail.re2.yahoo.com>
References: <802770.33034.qm@web52107.mail.re2.yahoo.com>
Message-ID: <368a5cd50903190131x76071323tf3f1fdf1c3a3eca4@mail.gmail.com>

The plan is to get 1.1.7 out of the door (early april) and then make a
version of that available for 3.x. There's also a preliminary port of 1.1.6
available from a third party; see the mailing list archives for pointers.

Cheers /F

On Mar 14, 2009 12:06 AM, "Cristi Constantin" <darkgl0w at yahoo.com> wrote:

Good day.

I just joined the mailing list.

I want to know if PIL will ever support Python 3. I am slowly migrating my
applications on Python 3 and i really need PIL.
It's not good to have half of my apps in Py 2.x and the rest in Py 3.x :)

Thank you in advance.

_______________________________________________
Image-SIG maillist  -  Image-SIG at python.org
http://mail.python.org/mailman/listinfo/image-sig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090319/689e4ec6/attachment.htm>

From fredrik at pythonware.com  Thu Mar 19 09:40:08 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 19 Mar 2009 09:40:08 +0100
Subject: [Image-SIG] png.save() and tEXt chunks
In-Reply-To: <ca816810903180643g54b6158dk60a87a46c7d7243d@mail.gmail.com>
References: <ca816810903180643g54b6158dk60a87a46c7d7243d@mail.gmail.com>
Message-ID: <368a5cd50903190140o7c44c489n8b7c0305b299a392@mail.gmail.com>

The current driver stores the tags in the info dictionary which is just a
collection of random stuff found by the format driver. Since that can
contain anything, including properties that depend on the image size or
contents, it's not preserved by internal operations, and not used by default
when saving an image.

The upcoming 1.1.7 release will improve things a bit; data from text chunks
are now mirrored in a separate dictionary, and I will add a text option to
the save method.

</F>

 On Mar 18, 2009 2:43 PM, "Kent Tenney" <ktenney at gmail.com> wrote:

Howdy,

I work with png's which have tEXt chunks containing metadata.
PIL discards them upon open .. save.

Might it be desirable for core save() to check for extra chunks
and preserve them if found?

Thanks,
Kent
_______________________________________________
Image-SIG maillist  -  Image-SIG at python.org
http://mail.python.org/mailman/listinfo/image-sig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090319/ec4fa21b/attachment.htm>

From ggpolo at gmail.com  Thu Mar 19 12:20:21 2009
From: ggpolo at gmail.com (Guilherme Polo)
Date: Thu, 19 Mar 2009 08:20:21 -0300
Subject: [Image-SIG] Python 3 support
In-Reply-To: <368a5cd50903190131x76071323tf3f1fdf1c3a3eca4@mail.gmail.com>
References: <802770.33034.qm@web52107.mail.re2.yahoo.com>
	<368a5cd50903190131x76071323tf3f1fdf1c3a3eca4@mail.gmail.com>
Message-ID: <ac2200130903190420r4f4802et844f2becb65a71a@mail.gmail.com>

2009/3/19 Fredrik Lundh <fredrik at pythonware.com>:
> The plan is to get 1.1.7 out of the door (early april) and then make a
> version of that available for 3.x. There's also a preliminary port of 1.1.6
> available from a third party; see the mailing list archives for pointers.
>

I hope to not cause even more confusion by replying here.

I believe the preliminary port to 1.1.6 Fredrik is talking about is
the one I provide. The link Cristi pointed also ends up pointing to my
port, where he said that cloning the git repo fails for him. So I
moved the git repo from my home to github today, you can clone it by
doing: git clone git://github.com/gpolo/pil-py3k.git and to see the
repo online you can visit: http://github.com/gpolo/pil-py3k

Hopefully this makes easier for you all to use, test and make it perfect (:))

> Cheers /F
>

Regards,

-- 
-- Guilherme H. Polo Goncalves

From darkgl0w at yahoo.com  Thu Mar 19 19:28:46 2009
From: darkgl0w at yahoo.com (Cristi Constantin)
Date: Thu, 19 Mar 2009 11:28:46 -0700 (PDT)
Subject: [Image-SIG] Python 3 support
Message-ID: <97281.95224.qm@web52103.mail.re2.yahoo.com>

Thank you very much for your answers.
I will wait for PIL 1.1.7 quietly. :)

Have a nice day.



      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090319/acfc506c/attachment.htm>

From charlie at begeistert.org  Thu Mar 19 20:50:47 2009
From: charlie at begeistert.org (Charlie Clark)
Date: Thu, 19 Mar 2009 20:50:47 +0100
Subject: [Image-SIG] Python 3 support
In-Reply-To: <97281.95224.qm@web52103.mail.re2.yahoo.com>
References: <97281.95224.qm@web52103.mail.re2.yahoo.com>
Message-ID: <538483ED-D576-4633-8B36-F0555C7920DF@begeistert.org>


Am 19.03.2009 um 19:28 schrieb Cristi Constantin:

> Thank you very much for your answers.
> I will wait for PIL 1.1.7 quietly. :)


While I think a lot of people would love to see a rapid migration to  
Python 3, it is a lot of work for many libraries (anything written in  
C) and, therefore, for the people who maintain them. In the meantime I  
think it is reasonable to expect users to target Python 2.6 which  
should encourage more Python 3.0 stuff without breaking important  
aspects of backward compatability for many projects.

Charlie
--
Charlie Clark
Helmholtzstr. 20
D?sseldorf
D- 40215
Tel: +49-211-938-5360
GSM: +49-178-782-6226




From ggpolo at gmail.com  Thu Mar 19 21:31:15 2009
From: ggpolo at gmail.com (Guilherme Polo)
Date: Thu, 19 Mar 2009 17:31:15 -0300
Subject: [Image-SIG] Python 3 support
In-Reply-To: <538483ED-D576-4633-8B36-F0555C7920DF@begeistert.org>
References: <97281.95224.qm@web52103.mail.re2.yahoo.com>
	<538483ED-D576-4633-8B36-F0555C7920DF@begeistert.org>
Message-ID: <ac2200130903191331p41c48d7cofb7a35946a688d65@mail.gmail.com>

On Thu, Mar 19, 2009 at 4:50 PM, Charlie Clark <charlie at begeistert.org> wrote:
>
> Am 19.03.2009 um 19:28 schrieb Cristi Constantin:
>
>> Thank you very much for your answers.
>> I will wait for PIL 1.1.7 quietly. :)
>
>
> While I think a lot of people would love to see a rapid migration to Python
> 3, it is a lot of work for many libraries (anything written in C)

I see the issue other way around, it is more troublesome to update
pure python code to 3.x than to port the C part.

> and,
> therefore, for the people who maintain them. In the meantime I think it is
> reasonable to expect users to target Python 2.6 which should encourage more
> Python 3.0 stuff without breaking important aspects of backward
> compatability for many projects.
>
> Charlie
> --
> Charlie Clark
> Helmholtzstr. 20
> D?sseldorf
> D- 40215
> Tel: +49-211-938-5360
> GSM: +49-178-782-6226
>
>
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>



-- 
-- Guilherme H. Polo Goncalves

From fetchinson at googlemail.com  Thu Mar 19 21:43:37 2009
From: fetchinson at googlemail.com (Daniel Fetchinson)
Date: Thu, 19 Mar 2009 13:43:37 -0700
Subject: [Image-SIG] Python 3 support
In-Reply-To: <ac2200130903190420r4f4802et844f2becb65a71a@mail.gmail.com>
References: <802770.33034.qm@web52107.mail.re2.yahoo.com>
	<368a5cd50903190131x76071323tf3f1fdf1c3a3eca4@mail.gmail.com>
	<ac2200130903190420r4f4802et844f2becb65a71a@mail.gmail.com>
Message-ID: <fbe2e2100903191343p72569166l1b55724aba142126@mail.gmail.com>

>> The plan is to get 1.1.7 out of the door (early april) and then make a
>> version of that available for 3.x. There's also a preliminary port of
>> 1.1.6
>> available from a third party; see the mailing list archives for pointers.
>>
>
> I hope to not cause even more confusion by replying here.
>
> I believe the preliminary port to 1.1.6 Fredrik is talking about is
> the one I provide. The link Cristi pointed also ends up pointing to my
> port, where he said that cloning the git repo fails for him. So I
> moved the git repo from my home to github today, you can clone it by
> doing: git clone git://github.com/gpolo/pil-py3k.git and to see the
> repo online you can visit: http://github.com/gpolo/pil-py3k

Actually, it turns out that the reason I could not check out your port
from the previous repo had to do with my laptop connection, proxies,
firewall, or what not. I could not check out from github either, but
when trying from another computer it all worked, so maybe your
previous repo would have worked either. Sorry about that.

Anyway, now I got the code and will start testing it over the weekend.

Cheers,
Daniel

> Hopefully this makes easier for you all to use, test and make it perfect
> (:))
>
>> Cheers /F


-- 
Psss, psss, put it down! - http://www.cafepress.com/putitdown

From michaelvantellingen at gmail.com  Tue Mar 24 09:54:07 2009
From: michaelvantellingen at gmail.com (Michael van Tellingen)
Date: Tue, 24 Mar 2009 09:54:07 +0100
Subject: [Image-SIG] [raclette] that JPEG CMYK thing, again
In-Reply-To: <368a5cd50903061626y68cd64f9u1615b8ab9564081a@mail.gmail.com>
References: <368a5cd50903061626y68cd64f9u1615b8ab9564081a@mail.gmail.com>
Message-ID: <f1e83f4f0903240154q1df4f3d4rfd21e8a55039dcbb@mail.gmail.com>

On Sat, Mar 7, 2009 at 01:26, Fredrik Lundh <fredrik at pythonware.com> wrote:

> Thanks to Tim and Charlie, I now have a set of nice Photoshop CMYK
> samples to play with.  The first result is a revamp of Kevin's CMYK
> patch; I'm now doing the inversion at the codec levels, both when
> reading and writing.  Also, the code *always* assumes Adobe behaviour,
> in both directions.  The current version of the plugin is here:
>
>
> http://bitbucket.org/effbot/pil-2009-raclette/src/e351bc0228c4/PIL/JpegImagePlugin.py
>
> I *think* this does the right thing, but I'm not 100% sure.  In other
> words, it would be helpful if those of you who've struggled with CMYK
> stuff can take this version for a test drive; it should work just fine
> together with a standard 1.1.6.
>

Hi Fredrik,

I've just tested my application with the PIL 1.1.7 tip and latest alpha and
still have problems with the conversion from CMYK jpeg to RGB.
The test images mailed earlier on this mailinglist (black background with
the text 'RED') somehow become yellow on my system (python 2.5 / osx
leopard).
(this worked fine with the earlier patches I used for 1.1.6 from Kevin
Cazabon)

The output of the pil setup summary:

    --------------------------------------------------------------------
    PIL 1.1.7a0 SETUP SUMMARY
    --------------------------------------------------------------------
    version       1.1.7a0
    platform      darwin 2.5.1 (r251:54863, Jan 13 2009, 10:26:13)
                  [GCC 4.0.1 (Apple Inc. build 5465)]
    --------------------------------------------------------------------
    --- TKINTER support available
    --- JPEG support available
    --- ZLIB (PNG/ZIP) support available
    --- FREETYPE2 support available
    --- LITTLECMS support available
    --------------------------------------------------------------------

I have a couple of photos with which i have the problems. I could mail you
these directly if you want.

Thanks,
Michael



>
> </F>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090324/67078354/attachment.htm>

From bryan at hideawaystudios.com  Tue Mar 24 17:15:18 2009
From: bryan at hideawaystudios.com (Bryan Jeal)
Date: Tue, 24 Mar 2009 10:15:18 -0600
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing JPEG
	file
Message-ID: <1175B076-BA92-40F4-AF2A-BA5AC98887DA@hideawaystudios.com>

Hello,

GOAL: Have a website built on the Django framework running under  
Windows Server 2003 (Apache with mod_python and MySQL) be able to  
upload pictures and then create a tinted thumbnail using PIL. I have  
all the code working and I have everything (python, PIL, django,  
apache and MySQL) setup and working on a linux machine; however, I  
HAVE to get it working on Windows.

PROBLEM: PIL with Jpeg support on Windows...

After much fighting with the Jpeg Library and PIL here is what I have  
accomplished:

Step 1: From Windows: Install python-2.5.4.msi with default settings,  
install Mingw32 and install Cygwin

Step 2: Start Cygwin and extract, './configure --enable-shared -- 
enable-static CC="gcc -mno-cygwin" CXX="gcc -mno-cygwin"', make, make - 
n install, make -n install-lib for jpeg-6b and extract, ./configure,  
make, make install for zlib

Step 3: From Windows: extract Imaging-1.1.6.tar.gz and edit setup.py  
to point to jpeg-6b and zlib library locations (ex. JPEG_ROOT = "C:\ 
\cygwin\\usr\\local\\src\\jpeg-6b\\.libs")

Step 4: From Windows: 'python setup.py build -c mingw32 install' the  
Imaging 1.1.6 package

Step 5: From Windows: 'python selftest.py' (NOTE: When building  
Imaging 1.1.6 jpeg and zlib support are OK) 57 tests passed for  
Imaging 1.1.6 package

Step 6: From Windows: 'python' then:
  >>> import Image
  >>> im = Image.open('c:/temp/testimage.jpg')
  >>> im.size[0] 									(result: 932)
  >>> tmpFile = open('C:/temp/temp_win.jpg', 'w')
  >>> im.save(tmpFile, 'JPEG', quality=90) 			(result: See Below)
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
   File "C:\Python25\lib\site-packages\PIL\Image.py", line 1405, in save
     save_handler(self, fp, filename)
   File "C:\Python25\lib\site-packages\PIL\JpegImagePlugin.py", line  
409, in _save
     ImageFile._save(im, fp, [("jpeg", (0,0)+im.size, 0, rawmode)])
   File "C:\Python25\lib\site-packages\PIL\ImageFile.py", line 491, in  
_save
     s = e.encode_to_file(fh, bufsize)
IOError: [Errno 0] Error

After finally getting over the _imaging C module is not installed  
(even though an _imaging.dll existed - Python2.5 seems to only like  
_imaging.pyd) and successfully opening and reading a JPEG (I had  
numerous other errors when I tried to get this working before) file I  
am so very lost. Please help!

Sincerely,

Bryan


From cannon.el at gmail.com  Tue Mar 24 18:13:25 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Tue, 24 Mar 2009 10:13:25 -0700
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <1175B076-BA92-40F4-AF2A-BA5AC98887DA@hideawaystudios.com>
References: <1175B076-BA92-40F4-AF2A-BA5AC98887DA@hideawaystudios.com>
Message-ID: <b986b290903241013x58de1a2bt328db4ca3f707fac@mail.gmail.com>

there are windows binary packages for both python and PIL available
(http://www.pythonware.com/products/pil/) and I have used these on XP
and Vista for years with 0 problem. I don't need Cygwin or Mingw32 or
anything else, just install the python and then install the binary. I
do jpeg all the time.

On Tue, Mar 24, 2009 at 9:15 AM, Bryan Jeal <bryan at hideawaystudios.com> wrote:
> Hello,
>
> GOAL: Have a website built on the Django framework running under Windows
> Server 2003 (Apache with mod_python and MySQL) be able to upload pictures
> and then create a tinted thumbnail using PIL. I have all the code working
> and I have everything (python, PIL, django, apache and MySQL) setup and
> working on a linux machine; however, I HAVE to get it working on Windows.
>
> PROBLEM: PIL with Jpeg support on Windows...
>
> After much fighting with the Jpeg Library and PIL here is what I have
> accomplished:
>
> Step 1: From Windows: Install python-2.5.4.msi with default settings,
> install Mingw32 and install Cygwin
>
> Step 2: Start Cygwin and extract, './configure --enable-shared
> --enable-static CC="gcc -mno-cygwin" CXX="gcc -mno-cygwin"', make, make -n
> install, make -n install-lib for jpeg-6b and extract, ./configure, make,
> make install for zlib
>
> Step 3: From Windows: extract Imaging-1.1.6.tar.gz and edit setup.py to
> point to jpeg-6b and zlib library locations (ex. JPEG_ROOT =
> "C:\\cygwin\\usr\\local\\src\\jpeg-6b\\.libs")
>
> Step 4: From Windows: 'python setup.py build -c mingw32 install' the Imaging
> 1.1.6 package
>
> Step 5: From Windows: 'python selftest.py' (NOTE: When building Imaging
> 1.1.6 jpeg and zlib support are OK) 57 tests passed for Imaging 1.1.6
> package
>
> Step 6: From Windows: 'python' then:
> ?>>> import Image
> ?>>> im = Image.open('c:/temp/testimage.jpg')
> ?>>> im.size[0]
> ? ? ? ? ? ? (result: 932)
> ?>>> tmpFile = open('C:/temp/temp_win.jpg', 'w')
> ?>>> im.save(tmpFile, 'JPEG', quality=90) ? ? ? ? ? ? ? ? ? ? ? (result: See
> Below)
> Traceback (most recent call last):
> ?File "<stdin>", line 1, in <module>
> ?File "C:\Python25\lib\site-packages\PIL\Image.py", line 1405, in save
> ? ?save_handler(self, fp, filename)
> ?File "C:\Python25\lib\site-packages\PIL\JpegImagePlugin.py", line 409, in
> _save
> ? ?ImageFile._save(im, fp, [("jpeg", (0,0)+im.size, 0, rawmode)])
> ?File "C:\Python25\lib\site-packages\PIL\ImageFile.py", line 491, in _save
> ? ?s = e.encode_to_file(fh, bufsize)
> IOError: [Errno 0] Error
>
> After finally getting over the _imaging C module is not installed (even
> though an _imaging.dll existed - Python2.5 seems to only like _imaging.pyd)
> and successfully opening and reading a JPEG (I had numerous other errors
> when I tried to get this working before) file I am so very lost. Please
> help!
>
> Sincerely,
>
> Bryan
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From Scott.Daniels at Acm.Org  Tue Mar 24 19:00:12 2009
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Tue, 24 Mar 2009 11:00:12 -0700
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <1175B076-BA92-40F4-AF2A-BA5AC98887DA@hideawaystudios.com>
References: <1175B076-BA92-40F4-AF2A-BA5AC98887DA@hideawaystudios.com>
Message-ID: <gqb6sp$2mb$1@ger.gmane.org>

Bryan Jeal wrote:
...
> After finally getting over the _imaging C module is not installed (even 
> though an _imaging.dll existed - Python2.5 seems to only like 
> _imaging.pyd) and successfully opening and reading a JPEG (I had 
> numerous other errors when I tried to get this working before) file I am 
> so very lost. Please help!
Laura & Edward Cannon told you the simplest way in general.
However, if you rename you _imaging.dll to be _imaging.pyd, your
troubles may also be over.  Because Microsoft has been trying to
increase their security by changing the rules about when dlls can
be loaded (no doubt due to some exciting virii), Python has
switched to using .pyd for the name of python dlls.

--Scott David Daniels
Scott.Daniels at Acm.Org



From fredrik at pythonware.com  Tue Mar 24 19:40:04 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 24 Mar 2009 19:40:04 +0100
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <gqb6sp$2mb$1@ger.gmane.org>
References: <1175B076-BA92-40F4-AF2A-BA5AC98887DA@hideawaystudios.com> 
	<gqb6sp$2mb$1@ger.gmane.org>
Message-ID: <368a5cd50903241140t5ba8de3cr6fb10c128540670c@mail.gmail.com>

Footnote: I'm pretty sure that the main reason for the renaming was
that we ended up with things like a binding called "zlib.dll" to a
low-level library named "zlib.dll".  See e.g. this excerpt from a
ten-year old copy of PC/readme.txt that I had lying around:

 /.../ The output file should be called
 "spam.dll" or "spam.pyd" (the latter is supported to avoid confusion with a
 system library "spam.dll" to which your module could be a Python interface)
 in Release mode, or spam_d.dll or spam_d.pyd in Debug mode.

Recent versions of Python no longer support importing DLL:s directly;
you can use imp.get_suffixes() to see what extensions a specific build
supports; here's Python 2.5:

>>> import imp
>>> imp.get_suffixes()
[('.pyd', 'rb', 3), ('.py', 'U', 1), ('.pyw', 'U', 1), ('.pyc', 'rb', 2)]
>>>

</F>

On Tue, Mar 24, 2009 at 7:00 PM, Scott David Daniels
<Scott.Daniels at acm.org> wrote:
> Bryan Jeal wrote:
> ...
>>
>> After finally getting over the _imaging C module is not installed (even
>> though an _imaging.dll existed - Python2.5 seems to only like _imaging.pyd)
>> and successfully opening and reading a JPEG (I had numerous other errors
>> when I tried to get this working before) file I am so very lost. Please
>> help!
>
> Laura & Edward Cannon told you the simplest way in general.
> However, if you rename you _imaging.dll to be _imaging.pyd, your
> troubles may also be over. ?Because Microsoft has been trying to
> increase their security by changing the rules about when dlls can
> be loaded (no doubt due to some exciting virii), Python has
> switched to using .pyd for the name of python dlls.
>
> --Scott David Daniels
> Scott.Daniels at Acm.Org
>
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From fredrik at pythonware.com  Wed Mar 25 01:09:53 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 25 Mar 2009 01:09:53 +0100
Subject: [Image-SIG] [raclette] that JPEG CMYK thing, again
In-Reply-To: <f1e83f4f0903240154q1df4f3d4rfd21e8a55039dcbb@mail.gmail.com>
References: <368a5cd50903061626y68cd64f9u1615b8ab9564081a@mail.gmail.com> 
	<f1e83f4f0903240154q1df4f3d4rfd21e8a55039dcbb@mail.gmail.com>
Message-ID: <368a5cd50903241709l345917e3l49b38e030ed6dd22@mail.gmail.com>

On Tue, Mar 24, 2009 at 9:54 AM, Michael van Tellingen
<michaelvantellingen at gmail.com> wrote:

> I've just tested my application with the PIL 1.1.7 tip and latest alpha and
> still have problems with the conversion from CMYK jpeg to RGB.
> The test images mailed earlier on this mailinglist (black background with
> the text 'RED') somehow become yellow on my system (python 2.5 / osx
> leopard).

Yeah, I completely messed this one up.  One crucial change didn't make
it into the library, and embarrasingly enough, I then went on to make
the JPEG regression test look for the broken behaviour...

Here's the missing patch:

$ hg diff libImaging/JpegDecode.c
diff -r f17d08c18ab3 libImaging/JpegDecode.c
--- a/libImaging/JpegDecode.c   Mon Mar 23 12:51:13 2009 +0100
+++ b/libImaging/JpegDecode.c   Wed Mar 25 01:04:59 2009 +0100
@@ -189,7 +189,8 @@
            context->cinfo.out_color_space = JCS_GRAYSCALE;
        else if (strcmp(context->rawmode, "RGB") == 0)
            context->cinfo.out_color_space = JCS_RGB;
-       else if (strcmp(context->rawmode, "CMYK") == 0)
+       else if (strcmp(context->rawmode, "CMYK") == 0 ||
+                 strcmp(context->rawmode, "CMYK;I") == 0)
            context->cinfo.out_color_space = JCS_CMYK;
        else if (strcmp(context->rawmode, "YCbCr") == 0)
            context->cinfo.out_color_space = JCS_YCbCr;

Thanks /F

From wgl0960 at hotmail.com  Wed Mar 25 05:11:52 2009
From: wgl0960 at hotmail.com (frankwong)
Date: Wed, 25 Mar 2009 04:11:52 +0000
Subject: [Image-SIG] I really need help for converting several still PNG
 images to an animated GIF..
Message-ID: <BAY119-W51E39DCD7AABCC04B3C24ABC900@phx.gbl>


I really need help for converting several still PNG images to an animated GIF..

 

Hello everyone,

       I need to convert several PNG images to an animated GIF.

       I searched a lot on the internet. And I found the script: gifmaker.py(http://svn.effbot.org/public/pil/Scripts/gifmaker.py)

 

       But I can not use it ..

 

       I drew several images for programming:(http://cid-a499a3cc7eaad0ae.skydrive.live.com/browse.aspx/HI%7C_PYTHON)

       There are 8 still PNG files and 1 animated GIF made by another software.

       Could anyone help me to convert these 8 still PNG images to animated GIF by PYTHON SCRIPT?

 

       Thank you.. I really need help.  Thank you. BTW, My English  is not good..

 

 

    

 

Frank wong..

_________________________________________________________________
MSN??????????????????MSN???
http://im.live.cn/safe/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090325/5fa5adca/attachment-0001.htm>

From bryan at hideawaystudios.com  Wed Mar 25 16:30:08 2009
From: bryan at hideawaystudios.com (Bryan Jeal)
Date: Wed, 25 Mar 2009 09:30:08 -0600
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
Message-ID: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com>

Here is an update:

1.
I removed Python and PIL from the system and reinstalled them (for PIL  
I used PIL-1.1.6.win32-py2.5.exe); however, it isn't working  
correctly...

The image I have loaded, resized and saved using PIL-1.1.6.win32- 
py2.5.exe. Has this error when viewing the JPEG:  Corrupt JPEG data: 3  
extraneous bytes before marker 0xdb

Images and code are here: http://test.hideawaystudios.com/testing/index.html

2.
I then removed Python and PIL again and installed python and built PIL  
from source. I then renamed _imaging.dll to _imaging.pyd and below is  
the error.

 >>> import _imaging
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: DLL load failed: The specified module could not be found.

Basically I am still lost... Any other ideas?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090325/9488107e/attachment.htm>

From fredrik at pythonware.com  Wed Mar 25 16:38:34 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 25 Mar 2009 16:38:34 +0100
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com>
References: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com>
Message-ID: <368a5cd50903250838m16372eaem21c751727f8f4fa4@mail.gmail.com>

2009/3/25 Bryan Jeal <bryan at hideawaystudios.com>:
> Here is an update:
>
> 1.
> I removed Python and PIL from the system and reinstalled them (for PIL I
> used PIL-1.1.6.win32-py2.5.exe); however, it isn't working correctly...

I'm not sure why you think you need to open the output file yourself,
but if you do, you need to tell Python that it's a binary file.
Changing

    tmpFile = open('C:/temp/temp_win.jpg', 'w')

to

    tmpFile = open('C:/temp/temp_win.jpg', 'wb')

or, easier, changing the save call to

    im.save("c/temp/temp_win.jpg", 'JPEG', quality=90)

will fix this.

</F>

From cannon.el at gmail.com  Wed Mar 25 16:45:44 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Wed, 25 Mar 2009 08:45:44 -0700
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <368a5cd50903250838m16372eaem21c751727f8f4fa4@mail.gmail.com>
References: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com>
	<368a5cd50903250838m16372eaem21c751727f8f4fa4@mail.gmail.com>
Message-ID: <b986b290903250845p23e97f40q65737d2a49ce43c@mail.gmail.com>

This is correct, windows makes a distinction between text and binary
files, and linux does not, I have had quite a bit of trouble in that
regard, especially as I switched from linux to windows.
Edward

On Wed, Mar 25, 2009 at 8:38 AM, Fredrik Lundh <fredrik at pythonware.com> wrote:
> 2009/3/25 Bryan Jeal <bryan at hideawaystudios.com>:
>> Here is an update:
>>
>> 1.
>> I removed Python and PIL from the system and reinstalled them (for PIL I
>> used PIL-1.1.6.win32-py2.5.exe); however, it isn't working correctly...
>
> I'm not sure why you think you need to open the output file yourself,
> but if you do, you need to tell Python that it's a binary file.
> Changing
>
> ? ?tmpFile = open('C:/temp/temp_win.jpg', 'w')
>
> to
>
> ? ?tmpFile = open('C:/temp/temp_win.jpg', 'wb')
>
> or, easier, changing the save call to
>
> ? ?im.save("c/temp/temp_win.jpg", 'JPEG', quality=90)
>
> will fix this.
>
> </F>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From fredrik at pythonware.com  Wed Mar 25 16:56:01 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 25 Mar 2009 16:56:01 +0100
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <368a5cd50903250838m16372eaem21c751727f8f4fa4@mail.gmail.com>
References: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com> 
	<368a5cd50903250838m16372eaem21c751727f8f4fa4@mail.gmail.com>
Message-ID: <368a5cd50903250856r12c16ce8l82f82cc0879aa197@mail.gmail.com>

> or, easier, changing the save call to
>
> ? ?im.save("c/temp/temp_win.jpg", 'JPEG', quality=90)

or, more likely to work:

    im.save("c:/temp/temp_win.jpg", 'JPEG', quality=90)

</F>

From Scott.Daniels at Acm.Org  Wed Mar 25 17:07:39 2009
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Wed, 25 Mar 2009 09:07:39 -0700
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com>
References: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com>
Message-ID: <gqdkls$854$1@ger.gmane.org>

Bryan Jeal wrote:
> ...
>  >>> import _imaging
> Traceback (most recent call last):
>  File "<stdin>", line 1, in <module>
> ImportError: DLL load failed: The specified module could not be found.
> 
> Basically I am still lost... Any other ideas?
When you run youtr python application with the command line:
     python -v prog.py ...
or even:
     python -vv prog.py ...
rather than your normal:
     python prog.py ...

You will get a lot of output on stdout showing what imports are tried
and what files are being looked for.  this output can be a source of
clues about where the issue lies.  This is only general advice about
how to debug import problems.  If you do ask again, please include in
your question the output of:

     import sys, platform
     print (sys.version)
     print (platform.platform())

as sometimes versions are quite important to the diagnosis.

--Scott David Daniels
Scott.Daniels at Acm.Org


From bryan at hideawaystudios.com  Wed Mar 25 17:34:24 2009
From: bryan at hideawaystudios.com (Bryan Jeal)
Date: Wed, 25 Mar 2009 10:34:24 -0600
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <gqdkls$854$1@ger.gmane.org>
References: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com>
	<gqdkls$854$1@ger.gmane.org>
Message-ID: <5D5D549D-106F-4852-BFC1-6E9B8F782178@hideawaystudios.com>

I would like to thank everyone for their help - you all are amazing!

It was a problem with not opening a file in binary mode - which as  
soon as I read that I smacked my head very hard and had a Homer moment.

I apologize for not including sys.version and platform.platform()  
information when I posted errors; however, I was running python -v for  
each of the tests I was running.

Everything is working perfectly and I hope I will get to help someone  
as you have helped me. Thank you very much for your time and knowledge.

Sincerely,

Bryan

From fredrik at pythonware.com  Wed Mar 25 18:14:41 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 25 Mar 2009 18:14:41 +0100
Subject: [Image-SIG] PIL with JPEG support on Windows: Error Writing
	JPEG file
In-Reply-To: <5D5D549D-106F-4852-BFC1-6E9B8F782178@hideawaystudios.com>
References: <9E47D490-6CF7-4D18-8E04-3124DA908955@hideawaystudios.com> 
	<gqdkls$854$1@ger.gmane.org>
	<5D5D549D-106F-4852-BFC1-6E9B8F782178@hideawaystudios.com>
Message-ID: <368a5cd50903251014y3301270cl8ec37e99ce8207ce@mail.gmail.com>

On Wed, Mar 25, 2009 at 5:34 PM, Bryan Jeal <bryan at hideawaystudios.com> wrote:
> I would like to thank everyone for their help - you all are amazing!
>
> It was a problem with not opening a file in binary mode - which as soon as I
> read that I smacked my head very hard and had a Homer moment.

I hope you realize that the reason some of us identify that kind of
mistakes very quickly is that we make them ourselves, all the time ;-)

</F>

From michaelvantellingen at gmail.com  Fri Mar 27 09:19:05 2009
From: michaelvantellingen at gmail.com (Michael van Tellingen)
Date: Fri, 27 Mar 2009 09:19:05 +0100
Subject: [Image-SIG] [raclette] that JPEG CMYK thing, again
In-Reply-To: <368a5cd50903241709l345917e3l49b38e030ed6dd22@mail.gmail.com>
References: <368a5cd50903061626y68cd64f9u1615b8ab9564081a@mail.gmail.com>
	<f1e83f4f0903240154q1df4f3d4rfd21e8a55039dcbb@mail.gmail.com>
	<368a5cd50903241709l345917e3l49b38e030ed6dd22@mail.gmail.com>
Message-ID: <f1e83f4f0903270119l3e2f275aj3d2b7956e4731689@mail.gmail.com>

Works perfectly now!

Thanks,
Michael


On Wed, Mar 25, 2009 at 01:09, Fredrik Lundh <fredrik at pythonware.com> wrote:

> On Tue, Mar 24, 2009 at 9:54 AM, Michael van Tellingen
> <michaelvantellingen at gmail.com> wrote:
>
> > I've just tested my application with the PIL 1.1.7 tip and latest alpha
> and
> > still have problems with the conversion from CMYK jpeg to RGB.
> > The test images mailed earlier on this mailinglist (black background with
> > the text 'RED') somehow become yellow on my system (python 2.5 / osx
> > leopard).
>
> Yeah, I completely messed this one up.  One crucial change didn't make
> it into the library, and embarrasingly enough, I then went on to make
> the JPEG regression test look for the broken behaviour...
>
> Here's the missing patch:
>
> $ hg diff libImaging/JpegDecode.c
> diff -r f17d08c18ab3 libImaging/JpegDecode.c
> --- a/libImaging/JpegDecode.c   Mon Mar 23 12:51:13 2009 +0100
> +++ b/libImaging/JpegDecode.c   Wed Mar 25 01:04:59 2009 +0100
> @@ -189,7 +189,8 @@
>            context->cinfo.out_color_space = JCS_GRAYSCALE;
>        else if (strcmp(context->rawmode, "RGB") == 0)
>            context->cinfo.out_color_space = JCS_RGB;
> -       else if (strcmp(context->rawmode, "CMYK") == 0)
> +       else if (strcmp(context->rawmode, "CMYK") == 0 ||
> +                 strcmp(context->rawmode, "CMYK;I") == 0)
>            context->cinfo.out_color_space = JCS_CMYK;
>        else if (strcmp(context->rawmode, "YCbCr") == 0)
>            context->cinfo.out_color_space = JCS_YCbCr;
>
> Thanks /F
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090327/d28ef33a/attachment.htm>

From fredrik at pythonware.com  Sat Mar 28 14:42:19 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 28 Mar 2009 14:42:19 +0100
Subject: [Image-SIG] ANN: PIL 1.1.7 alpha 2 (March 28, 2009)
Message-ID: <368a5cd50903280642r4357c87bl9536f15d0d54a2e7@mail.gmail.com>

PIL 1.1.7 alpha 2 for Python 2.X is now available.  Source code can be
downloaded using Subversion or Mercurial:

    http://svn.effbot.org/public/tags/pil-1.1.7a2-20090328/
    http://hg.effbot.org/pil-2009-raclette/src/pil-117a2-20090328/

Windows binaries for Python 2.4, 2.5, and 2.6 can be found here:

    http://effbot.org/downloads/#pil

I will post a standard tarball and Windows builds for 2.2 and 2.3
shortly.  I've done some basic testing with earlier versions; the kit
should build under 2.1, and will probably be possible to build also
under earlier versions with some light tweaking.

This release can be considered as "pretty much feature complete"; what
remains is some internal cleanup work, which probably won't affect the
public API (but I still reserve the right to tweak things, hence the
alpha label).  Many thanks to everyone who's helped testing alpha 2 in
various configurations.

</F>

From fredrik at pythonware.com  Sat Mar 28 17:36:27 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 28 Mar 2009 17:36:27 +0100
Subject: [Image-SIG] ANN: PIL 1.1.7 alpha 2 (March 28, 2009)
In-Reply-To: <368a5cd50903280642r4357c87bl9536f15d0d54a2e7@mail.gmail.com>
References: <368a5cd50903280642r4357c87bl9536f15d0d54a2e7@mail.gmail.com>
Message-ID: <368a5cd50903280936y707223a9lfacd36b36a3d29d8@mail.gmail.com>

On Sat, Mar 28, 2009 at 2:42 PM, Fredrik Lundh <fredrik at pythonware.com> wrote:
> PIL 1.1.7 alpha 2 for Python 2.X is now available. ?Source code can be
> downloaded using Subversion or Mercurial:
>
> ? ?http://svn.effbot.org/public/tags/pil-1.1.7a2-20090328/
> ? ?http://hg.effbot.org/pil-2009-raclette/src/pil-117a2-20090328/
>
> Windows binaries for Python 2.4, 2.5, and 2.6 can be found here:
>
> ? ?http://effbot.org/downloads/#pil

Binaries for 2.1 (!), 2.2, and 2.3 are now available from the same
location.  The a2 source kit doesn't build out of the box under 2.1;
if you need this, apply this patch:

    http://hg.effbot.org/pil-2009-raclette/changeset/526317446775/

</F>

From joeoettinger at gmail.com  Sat Mar 28 20:33:58 2009
From: joeoettinger at gmail.com (joe oettinger)
Date: Sat, 28 Mar 2009 15:33:58 -0400
Subject: [Image-SIG] PIL for Python3
Message-ID: <f93261ea0903281233g505b7545nb477c981361a74ae@mail.gmail.com>

Sure could use PIL for Python3. Anybody know if/when it'll be done?
I imagine there are a few other programmers interested. If there is an
burdensome number of e-mails like this, it might be useful to post the
answer on the image-sig web page.
Thanks in advance for any news.
Joe Oettinger
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20090328/e9fd094e/attachment.htm>

From fredrik at pythonware.com  Sat Mar 28 21:15:26 2009
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 28 Mar 2009 21:15:26 +0100
Subject: [Image-SIG] PIL for Python3
In-Reply-To: <f93261ea0903281233g505b7545nb477c981361a74ae@mail.gmail.com>
References: <f93261ea0903281233g505b7545nb477c981361a74ae@mail.gmail.com>
Message-ID: <368a5cd50903281315q47d9cdc6h418da058440af4ae@mail.gmail.com>

Checking recent posts can sometimes be helpful:

   http://mail.python.org/pipermail/image-sig/2009-March/005498.html

</F>

2009/3/28 joe oettinger <joeoettinger at gmail.com>:
> Sure could use PIL for Python3. Anybody know if/when it'll be done?
> I imagine there are a few other programmers interested. If there is an
> burdensome number of e-mails like this, it might be useful to post the
> answer on the image-sig web page.
> Thanks in advance for any news.
> Joe Oettinger
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>

From cannon.el at gmail.com  Sun Mar 29 05:15:14 2009
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Sat, 28 Mar 2009 20:15:14 -0700
Subject: [Image-SIG] autodetecting fonts
Message-ID: <b986b290903282015v74fe2d56r8037c631ede7039d@mail.gmail.com>

using PIL is there a good way to autodetect what fonts a user has installed?
This would be useful for applications that wanted to let a user select
a font, that way no assumptions have to be made about the fonts on a
particular computer and the user would have access to the widest range
of fonts possible.
Edward

From spe.stani.be at gmail.com  Sun Mar 29 13:58:03 2009
From: spe.stani.be at gmail.com (Stani)
Date: Sun, 29 Mar 2009 13:58:03 +0200
Subject: [Image-SIG] autodetecting fonts
In-Reply-To: <b986b290903282015v74fe2d56r8037c631ede7039d@mail.gmail.com>
References: <b986b290903282015v74fe2d56r8037c631ede7039d@mail.gmail.com>
Message-ID: <1238327883.14930.12.camel@blue>

Unfortunately this not possible at the moment as far as I know. For
Phatch I developed this module:
http://bazaar.launchpad.net/~stani/phatch/dev/annotate/head%
3A/phatch/core/lib//fonts.py

It is far from ideal, but it works.

You can see it in action when you run phatch
(http://photobatch.stani.be) and add a text action.

Stani

Op zaterdag 28-03-2009 om 20:15 uur [tijdzone -0700], schreef Laura &
Edward Cannon:
> using PIL is there a good way to autodetect what fonts a user has installed?
> This would be useful for applications that wanted to let a user select
> a font, that way no assumptions have to be made about the fonts on a
> particular computer and the user would have access to the widest range
> of fonts possible.
> Edward
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig