From benjaminrk at gmail.com  Sat Aug  1 19:00:34 2015
From: benjaminrk at gmail.com (MinRK)
Date: Sat, 1 Aug 2015 16:00:34 -0700
Subject: [IPython-dev] Issue with JupyterHub 0.2.0, ipython 3.2.1 and SSL
In-Reply-To: <CAOLi1KD2WkYicS92biDPRAPDHUQsjvs4pT47Yj-qyeYXRarnBQ@mail.gmail.com>
References: <CAOLi1KDmcQ0Pfoo702iveQLTWciRP_tB=SRVgYo-dRN_Oa1aTw@mail.gmail.com>
	<CAOvn4qiA8eRfGOV+BoL0FMmRrnP884uC9yk35wLhfAt6FJ8wjQ@mail.gmail.com>
	<CAOLi1KD2WkYicS92biDPRAPDHUQsjvs4pT47Yj-qyeYXRarnBQ@mail.gmail.com>
Message-ID: <CAHNn8BUaBRS9=kDG=bySmmzEBD8FfUkJGd1Mr=HEVvYOoO2ffw@mail.gmail.com>

Safari only refuses to open websocket connections with untrusted
certificates (e.g. self-signed certs). If you have a fully trusted
certificate, Safari works fine with HTTPS.

-MinRK

On Fri, Jul 31, 2015 at 2:27 PM, Francois Dion <francois.dion at gmail.com>
wrote:

> Didn't even think of that. Yes, I was using Safari. Just tried now and I'm
> glad to see it works in Chrome. Excellent way to finish the week.
>
> Thanks!
>
> On Fri, Jul 31, 2015 at 4:41 PM, Thomas Kluyver <takowl at gmail.com> wrote:
>
>> On 31 July 2015 at 12:46, Francois Dion <francois.dion at gmail.com> wrote:
>>
>>> But once I add a certificate and key and use https://fqdn, I get the
>>> login, I can login, but if I try to open a terminal, the new page stays
>>> empty, with just the jupyter logo. If I open a notebook, the notebook
>>> opens, but kernel communication doesn't happen... log shows kernel started.
>>>
>>
>> Are you testing in Safari? IIRC, Safari has trouble with websockets over
>> HTTPS, so if so, try it in Firefox or Chrome.
>>
>> _______________________________________________
>> IPython-dev mailing list
>> IPython-dev at scipy.org
>> http://mail.scipy.org/mailman/listinfo/ipython-dev
>>
>>
>
>
> --
> raspberry-python.blogspot.com - www.pyptug.org - www.3DFutureTech.info -
> @f_dion
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150801/ff12075b/attachment.html>

From kasturisurya at gmail.com  Sun Aug  2 15:10:50 2015
From: kasturisurya at gmail.com (Surya)
Date: Mon, 3 Aug 2015 00:40:50 +0530
Subject: [IPython-dev] nbviewer service in scipy central
Message-ID: <CAArdDCrO5ePeyvUwTRkxM5HStu7imUsfTxDUot=hKVbqAQYxSg@mail.gmail.com>

Hello there,

I am Surya, a contributor to SciPy Central Project -
https://github.com/scipy/SciPyCentral

There have been some discussions long time ago about migrating SciPy
Cookbooks to SciPy Central in ipynb format. I want to take this idea
further now.

I am wondering, if its possible for you guys to provide IPython Notebook
rendering as HTTP service for the application? It would help us to deeply
integrate feature.

Thanks
Surya
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150803/2a5bad5f/attachment.html>

From takowl at gmail.com  Sun Aug  2 16:45:10 2015
From: takowl at gmail.com (Thomas Kluyver)
Date: Sun, 2 Aug 2015 13:45:10 -0700
Subject: [IPython-dev] nbviewer service in scipy central
In-Reply-To: <CAArdDCrO5ePeyvUwTRkxM5HStu7imUsfTxDUot=hKVbqAQYxSg@mail.gmail.com>
References: <CAArdDCrO5ePeyvUwTRkxM5HStu7imUsfTxDUot=hKVbqAQYxSg@mail.gmail.com>
Message-ID: <CAOvn4qjNHU3sBx6SPDUU805pEnK7w6EOv=rat2K8ojH-zqu_UA@mail.gmail.com>

Hi Surya,

Thanks for working on this, I think notebooks would be a very natural
extension of SciPy Central.

On 2 August 2015 at 12:10, Surya <kasturisurya at gmail.com> wrote:

> I am wondering, if its possible for you guys to provide IPython Notebook
> rendering as HTTP service for the application? It would help us to deeply
> integrate feature.


On the technical side, yes, it's relatively easy to do notebook rendering
as an HTTP service. This is the basis of our own nbviewer site, and of
notebook rendering on Github.

Our web services are generously hosted for free by Rackspace. Kyle Kelley
will probably respond to this thread soon, and he can talk about what they
can offer for doing something like this.

Best wishes,
Thomas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150802/ce6b3be7/attachment.html>

From nick.bollweg at gmail.com  Sun Aug  2 23:15:25 2015
From: nick.bollweg at gmail.com (Nicholas Bollweg)
Date: Mon, 03 Aug 2015 03:15:25 +0000
Subject: [IPython-dev] nbviewer service in scipy central
In-Reply-To: <CAOvn4qjNHU3sBx6SPDUU805pEnK7w6EOv=rat2K8ojH-zqu_UA@mail.gmail.com>
References: <CAArdDCrO5ePeyvUwTRkxM5HStu7imUsfTxDUot=hKVbqAQYxSg@mail.gmail.com>
	<CAOvn4qjNHU3sBx6SPDUU805pEnK7w6EOv=rat2K8ojH-zqu_UA@mail.gmail.com>
Message-ID: <CACejjWzOJHKGqeZp1va_irOBPeDiwOge0a6v6XXxBc49G0xxbA@mail.gmail.com>

As has been pointed out before, nbviewer as such is a pretty thin wrapper
around nbconvert, with most of the decisions there coming from the
nbviewer.*.org site to handle working with a growing collection of
backends, and a fairly high number of burst users.

Where do you host your files? What more do you want from us? If you host
your files as ipynb, say this top result:
http://scipy-central.org/item/22/5/building-a-simple-interactive-2d-data-viewer-with-matplotlib.ipynb

That would just become:
GET
http://nbviewer.org/url/scipy-central.org/item/22/5/building-a-simple-interactive-2d-data-viewer-with-matplotlib.ipynb

You pull that, and do some regex stuff serve it back out. I am pretty sure
we won't throttle you, especially if you are doing some caching yourself:
django makes it pretty easy.

You could even put it in an iframe. We could potentially add some js/CSS
that might detect whether it was embedded if all you really want to do was
to hide our chrome.

If you are going to be doing much custom stuff beyond small regexen, and
you control the files you are rendering, it may actually be easier to
integrate with nbconvert than nbviewer. since you are already running
django, could you not just use something really simple like this flask
snippet:
https://gist.github.com/yoavram/5128702

Throw a little per-view caching on it, and you've got basically what
nbviewer does, without as many headaches. But do be careful, if people are
going to be logging in: it's basically not safe to run untrusted notebooks,
even/especially ones in "moderation queues" where the notebooks would have
javascript-based access to important data. Best practice would be to run
nbviewer (or your djnbviewer) on a wholly separate domain!

On Sun, Aug 2, 2015 at 4:45 PM Thomas Kluyver <takowl at gmail.com> wrote:

> Hi Surya,
>
> Thanks for working on this, I think notebooks would be a very natural
> extension of SciPy Central.
>
> On 2 August 2015 at 12:10, Surya <kasturisurya at gmail.com> wrote:
>
>> I am wondering, if its possible for you guys to provide IPython Notebook
>> rendering as HTTP service for the application? It would help us to deeply
>> integrate feature.
>
>
> On the technical side, yes, it's relatively easy to do notebook rendering
> as an HTTP service. This is the basis of our own nbviewer site, and of
> notebook rendering on Github.
>
> Our web services are generously hosted for free by Rackspace. Kyle Kelley
> will probably respond to this thread soon, and he can talk about what they
> can offer for doing something like this.
>
> Best wishes,
> Thomas
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150803/f0c782dd/attachment.html>

From mattwcraig at gmail.com  Sun Aug  2 23:30:33 2015
From: mattwcraig at gmail.com (Matt Craig)
Date: Sun, 2 Aug 2015 22:30:33 -0500
Subject: [IPython-dev] Where should bugfix PRs go,
 and how would I include a javascript fix in my package?
Message-ID: <CAKxiDYm4C8orqu0yh=8o_u71ZJbUH_MBvfaWw7zFAkueiGcFfg@mail.gmail.com>

Hi,

A couple of questions:

1. I've got a fix to a couple of issues that come up when
Accordion widgets are nested. I've read the contributing guide on
github and am wondering whether I should make the PR on master in
ipython/ipython (since the issue occurs in ipython 3.2) or in
ipython/ipywidgets (since that is the Future)?

2. In a nutshell, the fixes, in the javascript for AccordionView, ensure
that titles and show/hide events are not propagated to Accordions that are
children of an Accordion. I'd like to include this fix in the package I'm
using the Accordions in (reducer <https://github.com/mwcraig/reducer>). I'm
fine with creating a custom Accordion python class that is used instead
of the ipython Accordion if the user is using a version of ipytohn that
doesn't contain the fix in 1.

My question: How do I connect my python accordion to my modified version
of the javascript AccordionView?

To be clear, I'm already distributing this as a python package on pypi.
Also, I went to the advanced jupyter tutorial at SciPy 2015, so
if the answer is I should go back and read hose materials more closely,
just tell me :).

Matt Craig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150802/5cd61f88/attachment.html>

From kasturisurya at gmail.com  Mon Aug  3 12:30:19 2015
From: kasturisurya at gmail.com (Surya)
Date: Mon, 3 Aug 2015 22:00:19 +0530
Subject: [IPython-dev] nbviewer service in scipy central
In-Reply-To: <CACejjWzOJHKGqeZp1va_irOBPeDiwOge0a6v6XXxBc49G0xxbA@mail.gmail.com>
References: <CAArdDCrO5ePeyvUwTRkxM5HStu7imUsfTxDUot=hKVbqAQYxSg@mail.gmail.com>
	<CAOvn4qjNHU3sBx6SPDUU805pEnK7w6EOv=rat2K8ojH-zqu_UA@mail.gmail.com>
	<CACejjWzOJHKGqeZp1va_irOBPeDiwOge0a6v6XXxBc49G0xxbA@mail.gmail.com>
Message-ID: <CAArdDCpOdbVZuZmPn8giAwvvrTFcwLP--X2iEzz-vXN6SjxcPA@mail.gmail.com>

Hello,


On Mon, Aug 3, 2015 at 8:45 AM, Nicholas Bollweg <nick.bollweg at gmail.com>
wrote:

> As has been pointed out before, nbviewer as such is a pretty thin wrapper
> around nbconvert, with most of the decisions there coming from the
> nbviewer.*.org site to handle working with a growing collection of
> backends, and a fairly high number of burst users.
>
> Where do you host your files? What more do you want from us? If you host
> your files as ipynb, say this top result:
>
> http://scipy-central.org/item/22/5/building-a-simple-interactive-2d-data-viewer-with-matplotlib.ipynb
>  ?
>
>
That would just become:
> GET
> http://nbviewer.org/url/scipy-central.org/item/22/5/building-a-simple-interactive-2d-data-viewer-with-matplotlib.ipynb
>
> You pull that, and do some regex stuff serve it back out. I am pretty sure
> we won't throttle you, especially if you are doing some caching yourself:
> django makes it pretty easy.
>
> ?Yes, we'd like to host on SciPy Central itself. This sound sound good. I
will check with this implementation, and get back if there are any issues.
?


> You could even put it in an iframe. We could potentially add some js/CSS
> that might detect whether it was embedded if all you really want to do was
> to hide our chrome.
>
> If you are going to be doing much custom stuff beyond small regexen, and
> you control the files you are rendering, it may actually be easier to
> integrate with nbconvert than nbviewer. since you are already running
> django, could you not just use something really simple like this flask
> snippet:
> https://gist.github.com/yoavram/5128702
>
> Throw a little per-view caching on it, and you've got basically what
> nbviewer does, without as many headaches. But do be careful, if people are
> going to be logging in: it's basically not safe to run untrusted notebooks,
> even/especially ones in "moderation queues" where the notebooks would have
> javascript-based access to important data. Best practice would be to run
> nbviewer (or your djnbviewer) on a wholly separate domain!
>
> On Sun, Aug 2, 2015 at 4:45 PM Thomas Kluyver <takowl at gmail.com> wrote:
>
>> Hi Surya,
>>
>> Thanks for working on this, I think notebooks would be a very natural
>> extension of SciPy Central.
>>
>> On 2 August 2015 at 12:10, Surya <kasturisurya at gmail.com> wrote:
>>
>>> I am wondering, if its possible for you guys to provide IPython Notebook
>>> rendering as HTTP service for the application? It would help us to deeply
>>> integrate feature.
>>
>>
>> On the technical side, yes, it's relatively easy to do notebook rendering
>> as an HTTP service. This is the basis of our own nbviewer site, and of
>> notebook rendering on Github.
>>
>> Our web services are generously hosted for free by Rackspace. Kyle Kelley
>> will probably respond to this thread soon, and he can talk about what they
>> can offer for doing something like this.
>>
>> Best wishes,
>> Thomas
>> _______________________________________________
>> IPython-dev mailing list
>> IPython-dev at scipy.org
>> http://mail.scipy.org/mailman/listinfo/ipython-dev
>>
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
?Thank you guys.

Surya?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150803/b21c3410/attachment.html>

From jon.freder at gmail.com  Mon Aug  3 13:26:19 2015
From: jon.freder at gmail.com (Jonathan Frederic)
Date: Mon, 3 Aug 2015 10:26:19 -0700
Subject: [IPython-dev] Where should bugfix PRs go,
 and how would I include a javascript fix in my package?
In-Reply-To: <CAKxiDYm4C8orqu0yh=8o_u71ZJbUH_MBvfaWw7zFAkueiGcFfg@mail.gmail.com>
References: <CAKxiDYm4C8orqu0yh=8o_u71ZJbUH_MBvfaWw7zFAkueiGcFfg@mail.gmail.com>
Message-ID: <CAAoBLw0XWvsDxqh2zEQjSm5aXbQnpLqqMu7A1kre_hYiW27HOA@mail.gmail.com>

Hey Matt,

Nice to see you on the ML.  Please submit the PR to ipython/ipywidgets, if
it needs to be backported, we will handle that.

You're question about hooking up the JS, within the confines of a PYPI
package, is a good one.  I think I rushed over that material because of the
time constraint.  You'll need to use the nbextension API, either calling
out to it in your setup.py (see jdfreder/jupyter-pip) or in a python file
that's a part of your package.  The second option would make install a two
step process  i.e. `install.py`, installation would then be `pip install -e
. && python -m reducer.install`.

Once you get that part figured out, your JS should be available on the
page.  You can then subclass accordion using code like this:
```python
class Accordion(ipywidgets.Accordion):
        _view_module = Unicode("nbextensions/reducer/accordion", sync=True)
        _view_name = Unicode("AccordionView", sync=True)
```
where the above points to AccordionView in nbextensions/reducer/accordion.js

Cheers,
Jon

On Sun, Aug 2, 2015 at 8:30 PM, Matt Craig <mattwcraig at gmail.com> wrote:

> Hi,
>
> A couple of questions:
>
> 1. I've got a fix to a couple of issues that come up when
> Accordion widgets are nested. I've read the contributing guide on
> github and am wondering whether I should make the PR on master in
> ipython/ipython (since the issue occurs in ipython 3.2) or in
> ipython/ipywidgets (since that is the Future)?
>
> 2. In a nutshell, the fixes, in the javascript for AccordionView, ensure
> that titles and show/hide events are not propagated to Accordions that are
> children of an Accordion. I'd like to include this fix in the package I'm
> using the Accordions in (reducer <https://github.com/mwcraig/reducer>).
> I'm fine with creating a custom Accordion python class that is used instead
> of the ipython Accordion if the user is using a version of ipytohn that
> doesn't contain the fix in 1.
>
> My question: How do I connect my python accordion to my modified version
> of the javascript AccordionView?
>
> To be clear, I'm already distributing this as a python package on pypi.
> Also, I went to the advanced jupyter tutorial at SciPy 2015, so
> if the answer is I should go back and read hose materials more closely,
> just tell me :).
>
> Matt Craig
>
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150803/b5aa6cea/attachment.html>

From robert.dodier at gmail.com  Mon Aug  3 18:55:52 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Mon, 3 Aug 2015 15:55:52 -0700
Subject: [IPython-dev] announcing Maxima-Fishbowl,
	an IPython kernel for Maxima
Message-ID: <CAAsY_sRqFAQ-6X677nj6aPxwNJEHxWANCsvi+oBDdpa4qmMGog@mail.gmail.com>

Hi, just wanted to let you know, I've succeeded in creating a kernel
for the Maxima computer algebra system
(http://maxima.sourceforge.net), based on the Fishbowl project of
Frederic Peschanski. Many thanks to Frederic. The Github project for
Maxima-Fishbowl is: https://github.com/robert-dodier/maxima-fishbowl

There are rough spots and unimplemented stuff -- perhaps the most
important is that user interactions for the sign of some term (e.g.
cases for < 0 or > 0) are not handled. But basic functionality seems
to be working reliably as expected -- the IPython notebook is really a
pretty nice UI for Maxima, e.g.:
http://nbviewer.ipython.org/github/robert-dodier/maxima-fishbowl/blob/master/MaximaFishbowlExample.ipynb

Maybe you could add Maxima-Fishbowl to the list of IPython kernels for
other languages? Just a thought -- if not, it's really no problem.

Thanks a lot for your help, & best wishes.

Robert Dodier


From rgbkrk at gmail.com  Mon Aug  3 19:06:44 2015
From: rgbkrk at gmail.com (Kyle Kelley)
Date: Mon, 3 Aug 2015 18:06:44 -0500
Subject: [IPython-dev] announcing Maxima-Fishbowl,
	an IPython kernel for Maxima
In-Reply-To: <CAAsY_sRqFAQ-6X677nj6aPxwNJEHxWANCsvi+oBDdpa4qmMGog@mail.gmail.com>
References: <CAAsY_sRqFAQ-6X677nj6aPxwNJEHxWANCsvi+oBDdpa4qmMGog@mail.gmail.com>
Message-ID: <CA+tbMaWvjh2FdQEvGDs3PiYYyDM9E6O3yijsd0a7wxPsn8CP=A@mail.gmail.com>

Awesome!

The list of kernels at
https://github.com/ipython/ipython/wiki/IPython-kernels-for-other-languages
is actually a wiki, feel free to add it!

This makes me hope for a GAP kernel soon too. :D


On Mon, Aug 3, 2015 at 5:55 PM, Robert Dodier <robert.dodier at gmail.com>
wrote:

> Hi, just wanted to let you know, I've succeeded in creating a kernel
> for the Maxima computer algebra system
> (http://maxima.sourceforge.net), based on the Fishbowl project of
> Frederic Peschanski. Many thanks to Frederic. The Github project for
> Maxima-Fishbowl is: https://github.com/robert-dodier/maxima-fishbowl
>
> There are rough spots and unimplemented stuff -- perhaps the most
> important is that user interactions for the sign of some term (e.g.
> cases for < 0 or > 0) are not handled. But basic functionality seems
> to be working reliably as expected -- the IPython notebook is really a
> pretty nice UI for Maxima, e.g.:
>
> http://nbviewer.ipython.org/github/robert-dodier/maxima-fishbowl/blob/master/MaximaFishbowlExample.ipynb
>
> Maybe you could add Maxima-Fishbowl to the list of IPython kernels for
> other languages? Just a thought -- if not, it's really no problem.
>
> Thanks a lot for your help, & best wishes.
>
> Robert Dodier
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>



-- 
Kyle Kelley (@rgbkrk <https://twitter.com/rgbkrk>; lambdaops.com,
developer.rackspace.com)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150803/5227aff7/attachment.html>

From wes.turner at gmail.com  Mon Aug  3 19:48:27 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Mon, 3 Aug 2015 18:48:27 -0500
Subject: [IPython-dev] nbviewer service in scipy central
In-Reply-To: <CAArdDCpOdbVZuZmPn8giAwvvrTFcwLP--X2iEzz-vXN6SjxcPA@mail.gmail.com>
References: <CAArdDCrO5ePeyvUwTRkxM5HStu7imUsfTxDUot=hKVbqAQYxSg@mail.gmail.com>
	<CAOvn4qjNHU3sBx6SPDUU805pEnK7w6EOv=rat2K8ojH-zqu_UA@mail.gmail.com>
	<CACejjWzOJHKGqeZp1va_irOBPeDiwOge0a6v6XXxBc49G0xxbA@mail.gmail.com>
	<CAArdDCpOdbVZuZmPn8giAwvvrTFcwLP--X2iEzz-vXN6SjxcPA@mail.gmail.com>
Message-ID: <CACfEFw_EUM6nM_KH5Lbvj2fhs9Nu57w0RhQj3U8vZsEKc8ZzVQ@mail.gmail.com>

Jupyter Nbviewer dockerfile and docker compose yml:

* https://registry.hub.docker.com/u/jupyter/nbviewer/dockerfile/
* https://github.com/jupyter/nbviewer/blob/master/docker-compose.yml

Static hosting, manual build:

* https://pypi.python.org/pypi/runipy
* https://pypi.python.org/pypi/pgs
On Aug 3, 2015 11:30 AM, "Surya" <kasturisurya at gmail.com> wrote:

> Hello,
>
>
> On Mon, Aug 3, 2015 at 8:45 AM, Nicholas Bollweg <nick.bollweg at gmail.com>
> wrote:
>
>> As has been pointed out before, nbviewer as such is a pretty thin wrapper
>> around nbconvert, with most of the decisions there coming from the
>> nbviewer.*.org site to handle working with a growing collection of
>> backends, and a fairly high number of burst users.
>>
>> Where do you host your files? What more do you want from us? If you host
>> your files as ipynb, say this top result:
>>
>> http://scipy-central.org/item/22/5/building-a-simple-interactive-2d-data-viewer-with-matplotlib.ipynb
>>  ?
>>
>>
> That would just become:
>> GET
>> http://nbviewer.org/url/scipy-central.org/item/22/5/building-a-simple-interactive-2d-data-viewer-with-matplotlib.ipynb
>>
>> You pull that, and do some regex stuff serve it back out. I am pretty
>> sure we won't throttle you, especially if you are doing some caching
>> yourself: django makes it pretty easy.
>>
>> ?Yes, we'd like to host on SciPy Central itself. This sound sound good. I
> will check with this implementation, and get back if there are any issues.
> ?
>
>
>> You could even put it in an iframe. We could potentially add some js/CSS
>> that might detect whether it was embedded if all you really want to do was
>> to hide our chrome.
>>
>> If you are going to be doing much custom stuff beyond small regexen, and
>> you control the files you are rendering, it may actually be easier to
>> integrate with nbconvert than nbviewer. since you are already running
>> django, could you not just use something really simple like this flask
>> snippet:
>> https://gist.github.com/yoavram/5128702
>>
>> Throw a little per-view caching on it, and you've got basically what
>> nbviewer does, without as many headaches. But do be careful, if people are
>> going to be logging in: it's basically not safe to run untrusted notebooks,
>> even/especially ones in "moderation queues" where the notebooks would have
>> javascript-based access to important data. Best practice would be to run
>> nbviewer (or your djnbviewer) on a wholly separate domain!
>>
>> On Sun, Aug 2, 2015 at 4:45 PM Thomas Kluyver <takowl at gmail.com> wrote:
>>
>>> Hi Surya,
>>>
>>> Thanks for working on this, I think notebooks would be a very natural
>>> extension of SciPy Central.
>>>
>>> On 2 August 2015 at 12:10, Surya <kasturisurya at gmail.com> wrote:
>>>
>>>> I am wondering, if its possible for you guys to provide IPython
>>>> Notebook rendering as HTTP service for the application? It would help us to
>>>> deeply integrate feature.
>>>
>>>
>>> On the technical side, yes, it's relatively easy to do notebook
>>> rendering as an HTTP service. This is the basis of our own nbviewer site,
>>> and of notebook rendering on Github.
>>>
>>> Our web services are generously hosted for free by Rackspace. Kyle
>>> Kelley will probably respond to this thread soon, and he can talk about
>>> what they can offer for doing something like this.
>>>
>>> Best wishes,
>>> Thomas
>>> _______________________________________________
>>> IPython-dev mailing list
>>> IPython-dev at scipy.org
>>> http://mail.scipy.org/mailman/listinfo/ipython-dev
>>>
>>
>> _______________________________________________
>> IPython-dev mailing list
>> IPython-dev at scipy.org
>> http://mail.scipy.org/mailman/listinfo/ipython-dev
>>
>>
> ?Thank you guys.
>
> Surya?
>
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150803/89251ec7/attachment.html>

From mattwcraig at gmail.com  Mon Aug  3 20:02:15 2015
From: mattwcraig at gmail.com (Matt Craig)
Date: Mon, 3 Aug 2015 19:02:15 -0500
Subject: [IPython-dev] Where should bugfix PRs go,
 and how would I include a javascript fix in my package?
In-Reply-To: <CAAoBLw0XWvsDxqh2zEQjSm5aXbQnpLqqMu7A1kre_hYiW27HOA@mail.gmail.com>
References: <CAKxiDYm4C8orqu0yh=8o_u71ZJbUH_MBvfaWw7zFAkueiGcFfg@mail.gmail.com>
	<CAAoBLw0XWvsDxqh2zEQjSm5aXbQnpLqqMu7A1kre_hYiW27HOA@mail.gmail.com>
Message-ID: <CAKxiDY=br904Oi=3nZ75EmTTZEsH+rb0_EwCQ=Kp6_kzP9hdfA@mail.gmail.com>

Hi Jon,

Thanks -- the PR is in, and it looks like jupyter-pip does exactly what I
want!

Matt


On Mon, Aug 3, 2015 at 12:26 PM, Jonathan Frederic <jon.freder at gmail.com>
wrote:

> Hey Matt,
>
> Nice to see you on the ML.  Please submit the PR to ipython/ipywidgets, if
> it needs to be backported, we will handle that.
>
> You're question about hooking up the JS, within the confines of a PYPI
> package, is a good one.  I think I rushed over that material because of the
> time constraint.  You'll need to use the nbextension API, either calling
> out to it in your setup.py (see jdfreder/jupyter-pip) or in a python file
> that's a part of your package.  The second option would make install a two
> step process  i.e. `install.py`, installation would then be `pip install -e
> . && python -m reducer.install`.
>
> Once you get that part figured out, your JS should be available on the
> page.  You can then subclass accordion using code like this:
> ```python
> class Accordion(ipywidgets.Accordion):
>         _view_module = Unicode("nbextensions/reducer/accordion", sync=True)
>         _view_name = Unicode("AccordionView", sync=True)
> ```
> where the above points to AccordionView in
> nbextensions/reducer/accordion.js
>
> Cheers,
> Jon
>
> On Sun, Aug 2, 2015 at 8:30 PM, Matt Craig <mattwcraig at gmail.com> wrote:
>
>> Hi,
>>
>> A couple of questions:
>>
>> 1. I've got a fix to a couple of issues that come up when
>> Accordion widgets are nested. I've read the contributing guide on
>> github and am wondering whether I should make the PR on master in
>> ipython/ipython (since the issue occurs in ipython 3.2) or in
>> ipython/ipywidgets (since that is the Future)?
>>
>> 2. In a nutshell, the fixes, in the javascript for AccordionView, ensure
>> that titles and show/hide events are not propagated to Accordions that are
>> children of an Accordion. I'd like to include this fix in the package I'm
>> using the Accordions in (reducer <https://github.com/mwcraig/reducer>).
>> I'm fine with creating a custom Accordion python class that is used instead
>> of the ipython Accordion if the user is using a version of ipytohn that
>> doesn't contain the fix in 1.
>>
>> My question: How do I connect my python accordion to my modified version
>> of the javascript AccordionView?
>>
>> To be clear, I'm already distributing this as a python package on pypi.
>> Also, I went to the advanced jupyter tutorial at SciPy 2015, so
>> if the answer is I should go back and read hose materials more closely,
>> just tell me :).
>>
>> Matt Craig
>>
>>
>> _______________________________________________
>> IPython-dev mailing list
>> IPython-dev at scipy.org
>> http://mail.scipy.org/mailman/listinfo/ipython-dev
>>
>>
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150803/7aefc744/attachment.html>

From robert.dodier at gmail.com  Mon Aug  3 20:30:47 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Mon, 3 Aug 2015 17:30:47 -0700
Subject: [IPython-dev] announcing Maxima-Fishbowl,
	an IPython kernel for Maxima
In-Reply-To: <CA+tbMaWvjh2FdQEvGDs3PiYYyDM9E6O3yijsd0a7wxPsn8CP=A@mail.gmail.com>
References: <CAAsY_sRqFAQ-6X677nj6aPxwNJEHxWANCsvi+oBDdpa4qmMGog@mail.gmail.com>
	<CA+tbMaWvjh2FdQEvGDs3PiYYyDM9E6O3yijsd0a7wxPsn8CP=A@mail.gmail.com>
Message-ID: <CAAsY_sQ2nLxH=nLLrH9PvT6qr7Btz=bJ9hVruhhGME0sY+UE=w@mail.gmail.com>

Hi Kyle,

> The list of kernels at
> https://github.com/ipython/ipython/wiki/IPython-kernels-for-other-languages
> is actually a wiki, feel free to add it!


Oh, thanks for the tip. I have added it to the list.

best,

Robert


From fperez.net at gmail.com  Mon Aug  3 22:33:27 2015
From: fperez.net at gmail.com (Fernando Perez)
Date: Mon, 3 Aug 2015 19:33:27 -0700
Subject: [IPython-dev] announcing Maxima-Fishbowl,
	an IPython kernel for Maxima
In-Reply-To: <CAAsY_sQ2nLxH=nLLrH9PvT6qr7Btz=bJ9hVruhhGME0sY+UE=w@mail.gmail.com>
References: <CAAsY_sRqFAQ-6X677nj6aPxwNJEHxWANCsvi+oBDdpa4qmMGog@mail.gmail.com>
	<CA+tbMaWvjh2FdQEvGDs3PiYYyDM9E6O3yijsd0a7wxPsn8CP=A@mail.gmail.com>
	<CAAsY_sQ2nLxH=nLLrH9PvT6qr7Btz=bJ9hVruhhGME0sY+UE=w@mail.gmail.com>
Message-ID: <CAHAreOqHKPjR+jf1Ft9BsxXfR_5VpX_=FDDigtAxR07WgrVSyw@mail.gmail.com>

Robert, please do ping also the Sage list about this, both those folks and
the OpenDreamKit team will love to hear about this, since it will be right
up their alley.

Very exciting, thanks for sharing!

f

On Mon, Aug 3, 2015 at 5:30 PM, Robert Dodier <robert.dodier at gmail.com>
wrote:

> Hi Kyle,
>
> > The list of kernels at
> >
> https://github.com/ipython/ipython/wiki/IPython-kernels-for-other-languages
> > is actually a wiki, feel free to add it!
>
>
> Oh, thanks for the tip. I have added it to the list.
>
> best,
>
> Robert
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>



-- 
Fernando Perez (@fperez_org; http://fperez.org)
fperez.net-at-gmail: mailing lists only (I ignore this when swamped!)
fernando.perez-at-berkeley: contact me here for any direct mail
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150803/3e479153/attachment.html>

From glenn.caltech at gmail.com  Thu Aug  6 12:19:53 2015
From: glenn.caltech at gmail.com (G Jones)
Date: Thu, 6 Aug 2015 12:19:53 -0400
Subject: [IPython-dev] Customize IPython tree / dashboard with machine name
Message-ID: <CAGK_ABd7CDr-2=+Z7yZQiYm1i-g=wrBgeh0x=w7jL9sxXS+_xA@mail.gmail.com>

Hi,
I run ipython notebook servers on several different machines, and then set
up ssh tunnels from my personal machine to access the notebook servers in
my browser. The trouble is I have so many now that I am forgetting which
port number corresponds to which notebook server. So I'd like to add a bit
of text to the dashboard that says at least the machine name, and
preferably some other identifying info, in case I have two servers running
on the same machine. Is there an easy way to do this? I looked at the
generate_page_title function in html/tree/handlers.py, but it seems to be
hard coded to return either Home or the current directory path. Is there
something built in configurable? I'd like to avoid having to do anything
specialized as much as possible. I'm running ipython 3.1.0, but could
upgrade if this feature is available in a newer version.

Thanks,
Glenn
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150806/6e6e8f01/attachment.html>

From bussonniermatthias at gmail.com  Thu Aug  6 12:30:57 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Thu, 6 Aug 2015 09:30:57 -0700
Subject: [IPython-dev] Customize IPython tree / dashboard with machine
	name
In-Reply-To: <CAGK_ABd7CDr-2=+Z7yZQiYm1i-g=wrBgeh0x=w7jL9sxXS+_xA@mail.gmail.com>
References: <CAGK_ABd7CDr-2=+Z7yZQiYm1i-g=wrBgeh0x=w7jL9sxXS+_xA@mail.gmail.com>
Message-ID: <D2967458-CFAC-49A0-A043-EAD0B6A07CE3@gmail.com>


> On Aug 6, 2015, at 09:19, G Jones <glenn.caltech at gmail.com> wrote:
> 
> Hi,
> I run ipython notebook servers on several different machines, and then set up ssh tunnels from my personal machine to access the notebook servers in my browser. The trouble is I have so many now that I am forgetting which port number corresponds to which notebook server. So I'd like to add a bit of text to the dashboard that says at least the machine name, and preferably some other identifying info, in case I have two servers running on the same machine. Is there an easy way to do this? I looked at the generate_page_title function in html/tree/handlers.py, but it seems to be hard coded to return either Home or the current directory path. Is there something built in configurable? I'd like to avoid having to do anything specialized as much as possible. I'm running ipython 3.1.0, but could upgrade if this feature is available in a newer version.
> 


Out of curiosity, why are you running multiple servers ?

One server should be enough to browse all your hard drive and have access to many kernel ?

If you really need multiple server, it might be possible to host them behind a jupyterHub modified instance so that they are behind a unique port, 
but at different urls. 
Would that suit you ? 

-- 
M




From glenn.caltech at gmail.com  Thu Aug  6 12:34:54 2015
From: glenn.caltech at gmail.com (G Jones)
Date: Thu, 6 Aug 2015 12:34:54 -0400
Subject: [IPython-dev] Customize IPython tree / dashboard with machine
	name
In-Reply-To: <D2967458-CFAC-49A0-A043-EAD0B6A07CE3@gmail.com>
References: <CAGK_ABd7CDr-2=+Z7yZQiYm1i-g=wrBgeh0x=w7jL9sxXS+_xA@mail.gmail.com>
	<D2967458-CFAC-49A0-A043-EAD0B6A07CE3@gmail.com>
Message-ID: <CAGK_ABea8hAV1b00hbq8-pgpv7ToRR9btavFAgiDXcSuO6JQqQ@mail.gmail.com>

Hi,
I agree, the multiple server on one machine case is unnecessary; I did it
occasionally for working on two very different areas of research in
parallel, to just be able to instantly switch between the two.
Anyway, the main issue is that I have many machines on which I run a server
on, so accessing the whole hard drive doesn't help in that case.

Glenn

On Thu, Aug 6, 2015 at 12:30 PM, Matthias Bussonnier <
bussonniermatthias at gmail.com> wrote:

>
> > On Aug 6, 2015, at 09:19, G Jones <glenn.caltech at gmail.com> wrote:
> >
> > Hi,
> > I run ipython notebook servers on several different machines, and then
> set up ssh tunnels from my personal machine to access the notebook servers
> in my browser. The trouble is I have so many now that I am forgetting which
> port number corresponds to which notebook server. So I'd like to add a bit
> of text to the dashboard that says at least the machine name, and
> preferably some other identifying info, in case I have two servers running
> on the same machine. Is there an easy way to do this? I looked at the
> generate_page_title function in html/tree/handlers.py, but it seems to be
> hard coded to return either Home or the current directory path. Is there
> something built in configurable? I'd like to avoid having to do anything
> specialized as much as possible. I'm running ipython 3.1.0, but could
> upgrade if this feature is available in a newer version.
> >
>
>
> Out of curiosity, why are you running multiple servers ?
>
> One server should be enough to browse all your hard drive and have access
> to many kernel ?
>
> If you really need multiple server, it might be possible to host them
> behind a jupyterHub modified instance so that they are behind a unique port,
> but at different urls.
> Would that suit you ?
>
> --
> M
>
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150806/055aa7db/attachment.html>

From jsw at fnal.gov  Thu Aug  6 12:35:58 2015
From: jsw at fnal.gov (Jon Wilson)
Date: Thu, 6 Aug 2015 11:35:58 -0500
Subject: [IPython-dev] Customize IPython tree / dashboard with machine
 name
In-Reply-To: <D2967458-CFAC-49A0-A043-EAD0B6A07CE3@gmail.com>
References: <CAGK_ABd7CDr-2=+Z7yZQiYm1i-g=wrBgeh0x=w7jL9sxXS+_xA@mail.gmail.com>
	<D2967458-CFAC-49A0-A043-EAD0B6A07CE3@gmail.com>
Message-ID: <55C38CEE.2070204@fnal.gov>

I run multiple notebook servers because there are several different 
analysis machines that I use.  One here at A&M that is actually the head 
node of a cluster, three different machines at Stanford, one or more at 
Fermilab.  Different data and different computational capabilities are 
available in different places.  Some machines are available for analysis 
of data from my current experiment, and some only for data from my 
previous experiment.  So on and so forth.
Regards,
Jon

On 08/06/2015 11:30 AM, Matthias Bussonnier wrote:
>> On Aug 6, 2015, at 09:19, G Jones <glenn.caltech at gmail.com> wrote:
>>
>> Hi,
>> I run ipython notebook servers on several different machines, and then set up ssh tunnels from my personal machine to access the notebook servers in my browser. The trouble is I have so many now that I am forgetting which port number corresponds to which notebook server. So I'd like to add a bit of text to the dashboard that says at least the machine name, and preferably some other identifying info, in case I have two servers running on the same machine. Is there an easy way to do this? I looked at the generate_page_title function in html/tree/handlers.py, but it seems to be hard coded to return either Home or the current directory path. Is there something built in configurable? I'd like to avoid having to do anything specialized as much as possible. I'm running ipython 3.1.0, but could upgrade if this feature is available in a newer version.
>>
>
> Out of curiosity, why are you running multiple servers ?
>
> One server should be enough to browse all your hard drive and have access to many kernel ?
>
> If you really need multiple server, it might be possible to host them behind a jupyterHub modified instance so that they are behind a unique port,
> but at different urls.
> Would that suit you ?
>



From bussonniermatthias at gmail.com  Thu Aug  6 12:38:25 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Thu, 6 Aug 2015 09:38:25 -0700
Subject: [IPython-dev] Customize IPython tree / dashboard with machine
	name
In-Reply-To: <55C38CEE.2070204@fnal.gov>
References: <CAGK_ABd7CDr-2=+Z7yZQiYm1i-g=wrBgeh0x=w7jL9sxXS+_xA@mail.gmail.com>
	<D2967458-CFAC-49A0-A043-EAD0B6A07CE3@gmail.com>
	<55C38CEE.2070204@fnal.gov>
Message-ID: <7C6D992C-97F5-47FA-8BAE-601FBA5BF02F@gmail.com>

Sorry I missred  and missed the many machines. 
Jupyterhub should allow to redirect on many machines with some tweet, I?ll try to write a recap on how to customize the main page.

-- 
M


> On Aug 6, 2015, at 09:35, Jon Wilson <jsw at fnal.gov> wrote:
> 
> I run multiple notebook servers because there are several different 
> analysis machines that I use.  One here at A&M that is actually the head 
> node of a cluster, three different machines at Stanford, one or more at 
> Fermilab.  Different data and different computational capabilities are 
> available in different places.  Some machines are available for analysis 
> of data from my current experiment, and some only for data from my 
> previous experiment.  So on and so forth.
> Regards,
> Jon
> 
> On 08/06/2015 11:30 AM, Matthias Bussonnier wrote:
>>> On Aug 6, 2015, at 09:19, G Jones <glenn.caltech at gmail.com> wrote:
>>> 
>>> Hi,
>>> I run ipython notebook servers on several different machines, and then set up ssh tunnels from my personal machine to access the notebook servers in my browser. The trouble is I have so many now that I am forgetting which port number corresponds to which notebook server. So I'd like to add a bit of text to the dashboard that says at least the machine name, and preferably some other identifying info, in case I have two servers running on the same machine. Is there an easy way to do this? I looked at the generate_page_title function in html/tree/handlers.py, but it seems to be hard coded to return either Home or the current directory path. Is there something built in configurable? I'd like to avoid having to do anything specialized as much as possible. I'm running ipython 3.1.0, but could upgrade if this feature is available in a newer version.
>>> 
>> 
>> Out of curiosity, why are you running multiple servers ?
>> 
>> One server should be enough to browse all your hard drive and have access to many kernel ?
>> 
>> If you really need multiple server, it might be possible to host them behind a jupyterHub modified instance so that they are behind a unique port,
>> but at different urls.
>> Would that suit you ?
>> 
> 
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev



From bussonniermatthias at gmail.com  Thu Aug  6 12:49:25 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Thu, 6 Aug 2015 09:49:25 -0700
Subject: [IPython-dev] Customize IPython tree / dashboard with machine
	name
In-Reply-To: <55C38CEE.2070204@fnal.gov>
References: <CAGK_ABd7CDr-2=+Z7yZQiYm1i-g=wrBgeh0x=w7jL9sxXS+_xA@mail.gmail.com>
	<D2967458-CFAC-49A0-A043-EAD0B6A07CE3@gmail.com>
	<55C38CEE.2070204@fnal.gov>
Message-ID: <8EA5D40A-FD81-4EB5-8FEC-B88312E2FAFA@gmail.com>

So, 


Have look at this repo : 

https://github.com/jupyter/docker-demo-images <https://github.com/jupyter/docker-demo-images>


you see in particular that you can extend the page template: 

https://github.com/jupyter/docker-demo-images/blob/feea0967a3493a7ebe514f53084d84a7c8bfe88c/common/templates/page.html <https://github.com/jupyter/docker-demo-images/blob/feea0967a3493a7ebe514f53084d84a7c8bfe88c/common/templates/page.html>

(only line 1 and 16-20 of interest) that you can put somewhere on your hard drive:

```/path/to/your/templates/page.html

{% extends "templates/page.html" %}

{% block header_buttons %}
  <span id="login_widget">
    <a class="pull-right" style=?...?>YouText</a>
  </span>
{% endblock %}
```

and then extend the template search path as in here:

https://github.com/jupyter/docker-demo-images/blob/feea0967a3493a7ebe514f53084d84a7c8bfe88c/common/profile_default/ipython_notebook_config.py#L17 <https://github.com/jupyter/docker-demo-images/blob/feea0967a3493a7ebe514f53084d84a7c8bfe88c/common/profile_default/ipython_notebook_config.py#L17>

c.NotebookApp.extra_template_paths = [?/path/to/your/templates/?]

You can put text, image, or anything else that will apply to all your pages. 
You can also overwrite other template blocks. 

The JupyterHub case for 1 person many servers is an interesting one. 
I would bring it up as an issue on JupyterHub see if there are some abstractions we can
improve. 

-- 
M



> On Aug 6, 2015, at 09:35, Jon Wilson <jsw at fnal.gov> wrote:
> 
> I run multiple notebook servers because there are several different 
> analysis machines that I use.  One here at A&M that is actually the head 
> node of a cluster, three different machines at Stanford, one or more at 
> Fermilab.  Different data and different computational capabilities are 
> available in different places.  Some machines are available for analysis 
> of data from my current experiment, and some only for data from my 
> previous experiment.  So on and so forth.
> Regards,
> Jon
> 
> On 08/06/2015 11:30 AM, Matthias Bussonnier wrote:
>>> On Aug 6, 2015, at 09:19, G Jones <glenn.caltech at gmail.com> wrote:
>>> 
>>> Hi,
>>> I run ipython notebook servers on several different machines, and then set up ssh tunnels from my personal machine to access the notebook servers in my browser. The trouble is I have so many now that I am forgetting which port number corresponds to which notebook server. So I'd like to add a bit of text to the dashboard that says at least the machine name, and preferably some other identifying info, in case I have two servers running on the same machine. Is there an easy way to do this? I looked at the generate_page_title function in html/tree/handlers.py, but it seems to be hard coded to return either Home or the current directory path. Is there something built in configurable? I'd like to avoid having to do anything specialized as much as possible. I'm running ipython 3.1.0, but could upgrade if this feature is available in a newer version.
>>> 
>> 
>> Out of curiosity, why are you running multiple servers ?
>> 
>> One server should be enough to browse all your hard drive and have access to many kernel ?
>> 
>> If you really need multiple server, it might be possible to host them behind a jupyterHub modified instance so that they are behind a unique port,
>> but at different urls.
>> Would that suit you ?
>> 
> 
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150806/f368aa7f/attachment.html>

From kesmier84 at gmail.com  Sat Aug  8 08:42:14 2015
From: kesmier84 at gmail.com (Daniel Dean)
Date: Sat, 8 Aug 2015 08:42:14 -0400
Subject: [IPython-dev] IPython + R + SparkR
Message-ID: <501C9003-616B-40F2-BE33-3BA8EF70E2D0@gmail.com>

Hi,

We are interested in running an iPython notebook supporting SparkR. This is possible with PySpark, as illustrated by the well written article: 

http://ramhiser.com/2015/02/01/configuring-ipython-notebook-support-for-pyspark/

However, when trying a similar approach with R&SparkR, we found the shell does not load. Furthermore, we can manually load the SparkR library, but when performing any spark operations, the R kernel crashes. 

Is this type of functionality possible or would it require a custom kernel?

Regards,
Daniel



From robert.dodier at gmail.com  Sun Aug  9 17:14:44 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Sun, 9 Aug 2015 14:14:44 -0700
Subject: [IPython-dev] how to implement syntax highlighting for a new
	language
Message-ID: <CAAsY_sQyn6j9Bv6-UcGf3kpBO80yFOJBc+2OcNK8EgJvaJhBfA@mail.gmail.com>

Hi,

I have been working on Maxima-Fishbowl, an IPython kernel for the
computer algebra system Maxima. I would like to provide syntax
highlighting for Maxima code in the web browser and in documents
converted by nbconvert.

I gather that nbconvert uses pygments for syntax highlighting. It
looks like I could easily create a highlighter for Maxima by modifying
the one for MuPAD (a similar system). How can I tell IPython to use
the new highlighter?

I haven't been able to figure out how the browser does highlighting.
Fishbowl (Common Lisp kernel) has some custom Javascript which
apparently implements a highlighter based on CodeMirror. Is that the
usual way to go for new languages?

Thanks for your help, I appreciate it very much.

best,

Robert Dodier


From takowl at gmail.com  Sun Aug  9 17:44:57 2015
From: takowl at gmail.com (Thomas Kluyver)
Date: Sun, 9 Aug 2015 14:44:57 -0700
Subject: [IPython-dev] how to implement syntax highlighting for a new
	language
In-Reply-To: <CAAsY_sQyn6j9Bv6-UcGf3kpBO80yFOJBc+2OcNK8EgJvaJhBfA@mail.gmail.com>
References: <CAAsY_sQyn6j9Bv6-UcGf3kpBO80yFOJBc+2OcNK8EgJvaJhBfA@mail.gmail.com>
Message-ID: <CAOvn4qgF-qY+zsw0gv+EmKO57gALsE=JbbB=RjD1qVFgdCLzgg@mail.gmail.com>

On 9 August 2015 at 14:14, Robert Dodier <robert.dodier at gmail.com> wrote:

> I gather that nbconvert uses pygments for syntax highlighting. It
> looks like I could easily create a highlighter for Maxima by modifying
> the one for MuPAD (a similar system). How can I tell IPython to use
> the new highlighter?
>

Set up your lexer correctly as a pygments plugin:
http://pygments.org/docs/plugins/

And make sure that its name (or an alias) matches the language_info.name
you give in your kernel_info_reply.

If you can't use that same name for some reason, you can also specify
another language_info.pygments_lexer name in kernel_info_reply:
http://jupyter-client.readthedocs.org/en/latest/messaging.html#kernel-info


> I haven't been able to figure out how the browser does highlighting.
> Fishbowl (Common Lisp kernel) has some custom Javascript which
> apparently implements a highlighter based on CodeMirror. Is that the
> usual way to go for new languages?
>

Codemirror is the JS text editor we use for notebook cells. Many kernels
can just specify one of the modes codemirror already provides:
https://codemirror.net/mode/

If there isn't a suitable mode already in codemirror (it doesn't look like
there's one for Maxima), you'll need to provide one. You can do this in a
kernel.js file alongside kernel.json. (This doesn't seem to be documented
yet - I've made an issue for that:
https://github.com/jupyter/notebook/issues/265 ).

Thomas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150809/c807418f/attachment.html>

From ellisonbg at gmail.com  Sun Aug  9 22:36:31 2015
From: ellisonbg at gmail.com (Brian Granger)
Date: Sun, 9 Aug 2015 19:36:31 -0700
Subject: [IPython-dev] IPython + R + SparkR
In-Reply-To: <501C9003-616B-40F2-BE33-3BA8EF70E2D0@gmail.com>
References: <501C9003-616B-40F2-BE33-3BA8EF70E2D0@gmail.com>
Message-ID: <CAH4pYpS7_kNVWTrw_Hup5GM3KVK-Rr8BTe2bS8=x5AzHeKycUA@mail.gmail.com>

This should not require a custom kernel. Unlike PySpark, I think that
SparkR is installed as a regular R package. The only thing you should
have to do is set the SPARK_HOME. If there is a problem though, it
would probably be related to the event loop of the R kernel
conflicting with how R is talking to the JVM. I don't know anything
about that though. You might want to post this on the jupyter google
group as there are more people interested in jupyter+spark there.

On Sat, Aug 8, 2015 at 5:42 AM, Daniel Dean <kesmier84 at gmail.com> wrote:
> Hi,
>
> We are interested in running an iPython notebook supporting SparkR. This is possible with PySpark, as illustrated by the well written article:
>
> http://ramhiser.com/2015/02/01/configuring-ipython-notebook-support-for-pyspark/
>
> However, when trying a similar approach with R&SparkR, we found the shell does not load. Furthermore, we can manually load the SparkR library, but when performing any spark operations, the R kernel crashes.
>
> Is this type of functionality possible or would it require a custom kernel?
>
> Regards,
> Daniel
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev



-- 
Brian E. Granger
Cal Poly State University, San Luis Obispo
@ellisonbg on Twitter and GitHub
bgranger at calpoly.edu and ellisonbg at gmail.com


From benjaminrk at gmail.com  Tue Aug 11 13:43:03 2015
From: benjaminrk at gmail.com (MinRK)
Date: Tue, 11 Aug 2015 10:43:03 -0700
Subject: [IPython-dev] PySpark on PyPI discussion
Message-ID: <CAHNn8BWwuB=1VhZEj+KypxtCH23G5V+oQv1238EKwcpmzA70Tg@mail.gmail.com>

On the spark dev list
<http://apache-spark-developers-list.1001551.n3.nabble.com/PySpark-on-PyPi-td12626.html>,
there is a discussion about packaging and distributing PySpark, which may
be of interest to Jupyter and IPython users.

-MinRK
?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/7c674a19/attachment.html>

From nathan12343 at gmail.com  Tue Aug 11 14:38:03 2015
From: nathan12343 at gmail.com (Nathan Goldbaum)
Date: Tue, 11 Aug 2015 13:38:03 -0500
Subject: [IPython-dev] Restart notebook and run all cells button
Message-ID: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>

Hi all,

Has anyone put together a notebook extension that combines the "restart
kernel" and "run all cells" menu options into a single button or keyboard
shortcut? If so, can you point me to the appropriate repo?

Thanks for your help!

-Nathan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/15ee29df/attachment.html>

From bussonniermatthias at gmail.com  Tue Aug 11 15:12:26 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Tue, 11 Aug 2015 12:12:26 -0700
Subject: [IPython-dev] Restart notebook and run all cells button
In-Reply-To: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>
References: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>
Message-ID: <CANJQusVdoSLQvvGa_g-LAq0kN0X1hh2SLrvFuJL_e-hKywaLfA@mail.gmail.com>

Hi Nathan.

That was mostly what we did at scipy tutorial.
The extension can get some improvement, but you can see it here:

https://github.com/jupyter/scipy-advanced-tutorial/blob/master/Part1/extensions/hello-scipy-full.js

-- 
M


On Tue, Aug 11, 2015 at 11:38 AM, Nathan Goldbaum <nathan12343 at gmail.com> wrote:
> Hi all,
>
> Has anyone put together a notebook extension that combines the "restart
> kernel" and "run all cells" menu options into a single button or keyboard
> shortcut? If so, can you point me to the appropriate repo?
>
> Thanks for your help!
>
> -Nathan
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>


From wes.turner at gmail.com  Tue Aug 11 16:05:07 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Tue, 11 Aug 2015 15:05:07 -0500
Subject: [IPython-dev] Restart notebook and run all cells button
In-Reply-To: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>
References: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>
Message-ID: <CACfEFw-FiqkTvjjUoNDStGFRns08hZ6LbEWUOfN7Z+VZAOvJLA@mail.gmail.com>

* For the commandline (not js) runipy may be helpful:
https://github.com/paulgb/runipy

... from https://westurner.org/tools/#runipy

runipy? <https://westurner.org/tools/#runipy>
Src: git https://github.com/paulgb/runipy
PyPI: https://pypi.python.org/pypi/runipy

runipy runs Jupyter notebooks
<https://westurner.org/tools/#jupyter-notebook> from a Shell
<https://westurner.org/tools/#shells> commandline, generates HTML reports,
and can write errors to stderr.

Jupyter notebook <https://westurner.org/tools/#jupyter-notebook> *manual* test
review process:

# - run Jupyter Notebook server
!jupyter notebook
# - Browser
#     - navigate to / upload / drag and drop the notebook
        !web http://localhost:8888   # or https://
#     - (optional) click 'Cell' > 'All Output' > 'Clear'
#     - click 'Cell' > 'Run All'
#     - [wait] <Jupyter Kernel runs notebook>
#     - visually seek for the first ERRoring cell (scroll)
#     - review the notebook
        for (i, o) in notebook_cells:
            human.manually_review((i, o))
# - Compare the files on disk with the most recent commit (HEAD)
!git status && git diff
!git diff mynotebook.ipynb
# - Commit the changes
!git-add-commit "TST: mynotebook: tests for #123" ./mynotebook.ipynb

Jupyter notebook <https://westurner.org/tools/#jupyter-notebook> runipy
review process:

# - runipy the Jupyter notebook
!runipy mynotebook.ipynb
# - review stdout and stderr from runipy
# - review in browser (optional; recommended)
#     - navigate to the converted HTML
        !web ./mynotebook.ipynb.html
#     - visually seek for the first WEEoring cell (scroll)
#     - review the notebook
        for (i, o) in notebook_cells:
            human.manually_review((i, o))
# - Compare the files on disk with the most recent commit (HEAD)
!git status && git diff
!git diff mynotebook.ipynb*
# - Commit the changes
!git-add-commit "TST: mynotebook: tests for #123" ./mynotebook.ipynb*


   - An example of runipy usage in a Makefile
   <https://westurner.org/tools/#make>:
   https://github.com/westurner/notebooks/blob/gh-pages/Makefile


On Tue, Aug 11, 2015 at 1:38 PM, Nathan Goldbaum <nathan12343 at gmail.com>
wrote:

> Hi all,
>
> Has anyone put together a notebook extension that combines the "restart
> kernel" and "run all cells" menu options into a single button or keyboard
> shortcut? If so, can you point me to the appropriate repo?
>
> Thanks for your help!
>
> -Nathan
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/1f3a9367/attachment.html>

From wes.turner at gmail.com  Tue Aug 11 16:07:31 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Tue, 11 Aug 2015 15:07:31 -0500
Subject: [IPython-dev] Restart notebook and run all cells button
In-Reply-To: <CANJQusVdoSLQvvGa_g-LAq0kN0X1hh2SLrvFuJL_e-hKywaLfA@mail.gmail.com>
References: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>
	<CANJQusVdoSLQvvGa_g-LAq0kN0X1hh2SLrvFuJL_e-hKywaLfA@mail.gmail.com>
Message-ID: <CACfEFw8W5FTD5htg_GvAQYtDXEh0Pyq4iQ-UrynAQX9YLV4UzQ@mail.gmail.com>

On Tue, Aug 11, 2015 at 2:12 PM, Matthias Bussonnier <
bussonniermatthias at gmail.com> wrote:

> Hi Nathan.
>
> That was mostly what we did at scipy tutorial.
> The extension can get some improvement, but you can see it here:
>
>
> https://github.com/jupyter/scipy-advanced-tutorial/blob/master/Part1/extensions/hello-scipy-full.js
>
> --
> M
>
>
Thanks!


>
> On Tue, Aug 11, 2015 at 11:38 AM, Nathan Goldbaum <nathan12343 at gmail.com>
> wrote:
> > Hi all,
> >
> > Has anyone put together a notebook extension that combines the "restart
> > kernel" and "run all cells" menu options into a single button or keyboard
> > shortcut? If so, can you point me to the appropriate repo?
> >
> > Thanks for your help!
> >
> > -Nathan
> >
> > _______________________________________________
> > IPython-dev mailing list
> > IPython-dev at scipy.org
> > http://mail.scipy.org/mailman/listinfo/ipython-dev
> >
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/64a94899/attachment.html>

From nathan12343 at gmail.com  Tue Aug 11 16:14:44 2015
From: nathan12343 at gmail.com (Nathan Goldbaum)
Date: Tue, 11 Aug 2015 15:14:44 -0500
Subject: [IPython-dev] Restart notebook and run all cells button
In-Reply-To: <CACfEFw-FiqkTvjjUoNDStGFRns08hZ6LbEWUOfN7Z+VZAOvJLA@mail.gmail.com>
References: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>
	<CACfEFw-FiqkTvjjUoNDStGFRns08hZ6LbEWUOfN7Z+VZAOvJLA@mail.gmail.com>
Message-ID: <CAJXewOm=J3j=Rfba1vZvRB8-o3ZpXpPbWVho8WwJY3jm-MaARg@mail.gmail.com>

On Tue, Aug 11, 2015 at 3:05 PM, Wes Turner <wes.turner at gmail.com> wrote:

> * For the commandline (not js) runipy may be helpful:
> https://github.com/paulgb/runipy
>
>
Thanks for the suggestion. I've actually contributed to runipy in the past,
so I'm well aware of it. This isn't for the command line, it's because my
*most common* interaction with the notebook is to iterate on a problem by
restarting the kernel and running all the cells every time I change
something.  This avoids subtle nonlinear dependencies between cells and
satisfies my OCD need to have the cell execution numbers be sequential :)

It looks like Matthias' example is a good starting point. Thanks for the
link Matthias!


> ... from https://westurner.org/tools/#runipy
>
> runipy? <https://westurner.org/tools/#runipy>
> Src: git https://github.com/paulgb/runipy
> PyPI: https://pypi.python.org/pypi/runipy
>
> runipy runs Jupyter notebooks
> <https://westurner.org/tools/#jupyter-notebook> from a Shell
> <https://westurner.org/tools/#shells> commandline, generates HTML reports,
> and can write errors to stderr.
>
> Jupyter notebook <https://westurner.org/tools/#jupyter-notebook> *manual* test
> review process:
>
> # - run Jupyter Notebook server
> !jupyter notebook
> # - Browser
> #     - navigate to / upload / drag and drop the notebook
>         !web http://localhost:8888   # or https://
> #     - (optional) click 'Cell' > 'All Output' > 'Clear'
> #     - click 'Cell' > 'Run All'
> #     - [wait] <Jupyter Kernel runs notebook>
> #     - visually seek for the first ERRoring cell (scroll)
> #     - review the notebook
>         for (i, o) in notebook_cells:
>             human.manually_review((i, o))
> # - Compare the files on disk with the most recent commit (HEAD)
> !git status && git diff
> !git diff mynotebook.ipynb
> # - Commit the changes
> !git-add-commit "TST: mynotebook: tests for #123" ./mynotebook.ipynb
>
> Jupyter notebook <https://westurner.org/tools/#jupyter-notebook> runipy
> review process:
>
> # - runipy the Jupyter notebook
> !runipy mynotebook.ipynb
> # - review stdout and stderr from runipy
> # - review in browser (optional; recommended)
> #     - navigate to the converted HTML
>         !web ./mynotebook.ipynb.html
> #     - visually seek for the first WEEoring cell (scroll)
> #     - review the notebook
>         for (i, o) in notebook_cells:
>             human.manually_review((i, o))
> # - Compare the files on disk with the most recent commit (HEAD)
> !git status && git diff
> !git diff mynotebook.ipynb*
> # - Commit the changes
> !git-add-commit "TST: mynotebook: tests for #123" ./mynotebook.ipynb*
>
>
>    - An example of runipy usage in a Makefile
>    <https://westurner.org/tools/#make>:
>    https://github.com/westurner/notebooks/blob/gh-pages/Makefile
>
>
> On Tue, Aug 11, 2015 at 1:38 PM, Nathan Goldbaum <nathan12343 at gmail.com>
> wrote:
>
>> Hi all,
>>
>> Has anyone put together a notebook extension that combines the "restart
>> kernel" and "run all cells" menu options into a single button or keyboard
>> shortcut? If so, can you point me to the appropriate repo?
>>
>> Thanks for your help!
>>
>> -Nathan
>>
>> _______________________________________________
>> IPython-dev mailing list
>> IPython-dev at scipy.org
>> http://mail.scipy.org/mailman/listinfo/ipython-dev
>>
>>
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/ab5bfc2c/attachment.html>

From wes.turner at gmail.com  Tue Aug 11 16:56:44 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Tue, 11 Aug 2015 15:56:44 -0500
Subject: [IPython-dev] PySpark on PyPI discussion
In-Reply-To: <CAHNn8BWwuB=1VhZEj+KypxtCH23G5V+oQv1238EKwcpmzA70Tg@mail.gmail.com>
References: <CAHNn8BWwuB=1VhZEj+KypxtCH23G5V+oQv1238EKwcpmzA70Tg@mail.gmail.com>
Message-ID: <CACfEFw8ij_quURw+b5VVJe7cp3v+B7iekrNf+NeftO-i4i9JYg@mail.gmail.com>

On Tue, Aug 11, 2015 at 12:43 PM, MinRK <benjaminrk at gmail.com> wrote:

> On the spark dev list
> <http://apache-spark-developers-list.1001551.n3.nabble.com/PySpark-on-PyPi-td12626.html>,
> there is a discussion about packaging and distributing PySpark, which may
> be of interest to Jupyter and IPython users.
>
> -MinRK
>

Good call! CC'ing here because "This post has NOT been accepted by the
mailing list yet.":

$ conda skeleton pypi pyspark
# update git_tag and git_uri
# add test commands (import pyspark; import pyspark.[...])

Docs for building conda packages for multiple operating systems and
interpreters from PyPi packages:

*
http://www.pydanny.com/building-conda-packages-for-multiple-operating-systems.html
* https://github.com/audreyr/cookiecutter/issues/232



> ?
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/42e7cd0c/attachment.html>

From wes.turner at gmail.com  Tue Aug 11 17:04:17 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Tue, 11 Aug 2015 16:04:17 -0500
Subject: [IPython-dev] Restart notebook and run all cells button
In-Reply-To: <CACfEFw8W5FTD5htg_GvAQYtDXEh0Pyq4iQ-UrynAQX9YLV4UzQ@mail.gmail.com>
References: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>
	<CANJQusVdoSLQvvGa_g-LAq0kN0X1hh2SLrvFuJL_e-hKywaLfA@mail.gmail.com>
	<CACfEFw8W5FTD5htg_GvAQYtDXEh0Pyq4iQ-UrynAQX9YLV4UzQ@mail.gmail.com>
Message-ID: <CACfEFw-Fd4+x9otGNUyk3_P0LcHH-+-iOxq1pZV5CixveB5=+Q@mail.gmail.com>

On Tue, Aug 11, 2015 at 3:07 PM, Wes Turner <wes.turner at gmail.com> wrote:

>
>
> On Tue, Aug 11, 2015 at 2:12 PM, Matthias Bussonnier <
> bussonniermatthias at gmail.com> wrote:
>
>> Hi Nathan.
>>
>> That was mostly what we did at scipy tutorial.
>> The extension can get some improvement, but you can see it here:
>>
>>
>> https://github.com/jupyter/scipy-advanced-tutorial/blob/master/Part1/extensions/hello-scipy-full.js
>>
>> --
>> M
>>
>>
> Thanks!
>

How would I make this IPython/Jupyter extension installable as a package,
from pypi, or with pip, which does HTTPS correctly, with a version string?

- [ ] ENH: ipyextension2ipyextensionpackage (e.g. cookiecutter)


>
>
>>
>> On Tue, Aug 11, 2015 at 11:38 AM, Nathan Goldbaum <nathan12343 at gmail.com>
>> wrote:
>> > Hi all,
>> >
>> > Has anyone put together a notebook extension that combines the "restart
>> > kernel" and "run all cells" menu options into a single button or
>> keyboard
>> > shortcut? If so, can you point me to the appropriate repo?
>> >
>> > Thanks for your help!
>> >
>> > -Nathan
>> >
>> > _______________________________________________
>> > IPython-dev mailing list
>> > IPython-dev at scipy.org
>> > http://mail.scipy.org/mailman/listinfo/ipython-dev
>> >
>> _______________________________________________
>> IPython-dev mailing list
>> IPython-dev at scipy.org
>> http://mail.scipy.org/mailman/listinfo/ipython-dev
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/ed2ca86a/attachment.html>

From wes.turner at gmail.com  Tue Aug 11 17:06:14 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Tue, 11 Aug 2015 16:06:14 -0500
Subject: [IPython-dev] Restart notebook and run all cells button
In-Reply-To: <CACfEFw-Fd4+x9otGNUyk3_P0LcHH-+-iOxq1pZV5CixveB5=+Q@mail.gmail.com>
References: <CAJXewOnafJ2mRL2DnZkrS61bAg=fg1CJBb7MN2jnaOt_i+bYLw@mail.gmail.com>
	<CANJQusVdoSLQvvGa_g-LAq0kN0X1hh2SLrvFuJL_e-hKywaLfA@mail.gmail.com>
	<CACfEFw8W5FTD5htg_GvAQYtDXEh0Pyq4iQ-UrynAQX9YLV4UzQ@mail.gmail.com>
	<CACfEFw-Fd4+x9otGNUyk3_P0LcHH-+-iOxq1pZV5CixveB5=+Q@mail.gmail.com>
Message-ID: <CACfEFw_a3fNLYypABW86DzHqbz7D85aY3mEo=Uu=EhS=yDw4kw@mail.gmail.com>

On Tue, Aug 11, 2015 at 4:04 PM, Wes Turner <wes.turner at gmail.com> wrote:

>
>
> On Tue, Aug 11, 2015 at 3:07 PM, Wes Turner <wes.turner at gmail.com> wrote:
>
>>
>>
>> On Tue, Aug 11, 2015 at 2:12 PM, Matthias Bussonnier <
>> bussonniermatthias at gmail.com> wrote:
>>
>>> Hi Nathan.
>>>
>>> That was mostly what we did at scipy tutorial.
>>> The extension can get some improvement, but you can see it here:
>>>
>>>
>>> https://github.com/jupyter/scipy-advanced-tutorial/blob/master/Part1/extensions/hello-scipy-full.js
>>>
>>> --
>>> M
>>>
>>>
>> Thanks!
>>
>
> How would I make this IPython/Jupyter extension installable as a package,
> from pypi, or with pip, which does HTTPS correctly, with a version string?
>
> - [ ] ENH: ipyextension2ipyextensionpackage (e.g. cookiecutter)
>

https://github.com/jdfreder/jupyter-pip "Allows Jupyter notebook extension
writers to make their extension pip installable!"


>
>
>>
>>
>>>
>>> On Tue, Aug 11, 2015 at 11:38 AM, Nathan Goldbaum <nathan12343 at gmail.com>
>>> wrote:
>>> > Hi all,
>>> >
>>> > Has anyone put together a notebook extension that combines the "restart
>>> > kernel" and "run all cells" menu options into a single button or
>>> keyboard
>>> > shortcut? If so, can you point me to the appropriate repo?
>>> >
>>> > Thanks for your help!
>>> >
>>> > -Nathan
>>> >
>>> > _______________________________________________
>>> > IPython-dev mailing list
>>> > IPython-dev at scipy.org
>>> > http://mail.scipy.org/mailman/listinfo/ipython-dev
>>> >
>>> _______________________________________________
>>> IPython-dev mailing list
>>> IPython-dev at scipy.org
>>> http://mail.scipy.org/mailman/listinfo/ipython-dev
>>>
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/29ed4711/attachment.html>

From takowl at gmail.com  Tue Aug 11 18:47:40 2015
From: takowl at gmail.com (Thomas Kluyver)
Date: Tue, 11 Aug 2015 15:47:40 -0700
Subject: [IPython-dev] Script to run 2to3 on notebooks
Message-ID: <CAOvn4qi76K6AVAeRAv0R016zkyzXmC+fT-CCrW1p0a7yYM3UCA@mail.gmail.com>

Fernando asked me to put together a script that will run 2to3 on notebooks.
So here it is:

https://gist.github.com/takluyver/c8839593c615bb2f6e80

If you want to make it do more things, let me know, and I'll move it to a
proper git repository to facilitate collaboration. For now, it's just a 50
line script.

Thanks,
Thomas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/ff056472/attachment.html>

From benjaminrk at gmail.com  Tue Aug 11 22:03:04 2015
From: benjaminrk at gmail.com (MinRK)
Date: Tue, 11 Aug 2015 19:03:04 -0700
Subject: [IPython-dev] [ANN] Released IPython 4.0, completing The Big Split
Message-ID: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>

We?ve just released IPython 4.0, completing the transition of
language-agnostic pieces of IPython to their new homes on Jupyter. This is
primarily a reorganization of the code into the new repos, with a few bugs
fixed along the way.

To install the new jupyter everything, including IPython, the notebook,
qtconsole, etc.:

pip install --upgrade jupyter

and start the notebook with:

jupyter notebook

or get just the IPython terminal, without any of the Jupyter fun:

pip install --upgrade ipython

There should be shims for keep old commands and imports working during the
transition, so let us know if you find places where that?s not the case.

For more info, see the announcement
<http://blog.jupyter.org/2015/08/12/first-release-of-jupyter/>, and
the new jupyter
docs <https://jupyter.readthedocs.org/en/latest/>.

Thanks everyone!

-The IPython and Jupyter Teams
?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/54d9d313/attachment.html>

From fperez.net at gmail.com  Tue Aug 11 22:07:56 2015
From: fperez.net at gmail.com (Fernando Perez)
Date: Tue, 11 Aug 2015 19:07:56 -0700
Subject: [IPython-dev] [ANN] Released IPython 4.0,
	completing The Big Split
In-Reply-To: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
References: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
Message-ID: <CAHAreOpCAAm=3Fk9xSbZDWxBMaFq0v6sLhnWOiLF=zkAOOELkQ@mail.gmail.com>

Thanks so much to everyone who's put a crazy amount of work into this over
the last few months!

Cheers

f

On Tue, Aug 11, 2015 at 7:03 PM, MinRK <benjaminrk at gmail.com> wrote:

> We?ve just released IPython 4.0, completing the transition of
> language-agnostic pieces of IPython to their new homes on Jupyter. This is
> primarily a reorganization of the code into the new repos, with a few bugs
> fixed along the way.
>
> To install the new jupyter everything, including IPython, the notebook,
> qtconsole, etc.:
>
> pip install --upgrade jupyter
>
> and start the notebook with:
>
> jupyter notebook
>
> or get just the IPython terminal, without any of the Jupyter fun:
>
> pip install --upgrade ipython
>
> There should be shims for keep old commands and imports working during the
> transition, so let us know if you find places where that?s not the case.
>
> For more info, see the announcement
> <http://blog.jupyter.org/2015/08/12/first-release-of-jupyter/>, and the
> new jupyter docs <https://jupyter.readthedocs.org/en/latest/>.
>
> Thanks everyone!
>
> -The IPython and Jupyter Teams
> ?
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>


-- 
Fernando Perez (@fperez_org; http://fperez.org)
fperez.net-at-gmail: mailing lists only (I ignore this when swamped!)
fernando.perez-at-berkeley: contact me here for any direct mail
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150811/0423c750/attachment.html>

From DavidAnthonyPowell+python at gmail.com  Wed Aug 12 00:30:38 2015
From: DavidAnthonyPowell+python at gmail.com (David Powell)
Date: Wed, 12 Aug 2015 14:30:38 +1000
Subject: [IPython-dev] [ANN] Released IPython 4.0,
	completing The Big Split
In-Reply-To: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
References: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
Message-ID: <CAGXzvieZQBtV3YxRYVT0m1Tombq53_zHtYTJWYZs__wUyprSQQ@mail.gmail.com>

Great news, and thanks for the hard work.

I just used pip to upgrade my Windows conda based installation. One thing I
noticed is that all of the jupyter scripts are python files without
extensions, rather than windows .exe files. So "jupyter notebook" does not
work from the command line. However "ipython notebook" works just fine, as
there is a new ipython.exe in my scripts folder.

Should the installer be creating this .exe file? If it's a bug, to which
repo(s) should it be reported? It's not so clear after the big split.

regards
David


On 12 August 2015 at 12:03, MinRK <benjaminrk at gmail.com> wrote:

> We?ve just released IPython 4.0, completing the transition of
> language-agnostic pieces of IPython to their new homes on Jupyter. This is
> primarily a reorganization of the code into the new repos, with a few bugs
> fixed along the way.
>
> To install the new jupyter everything, including IPython, the notebook,
> qtconsole, etc.:
>
> pip install --upgrade jupyter
>
> and start the notebook with:
>
> jupyter notebook
>
> or get just the IPython terminal, without any of the Jupyter fun:
>
> pip install --upgrade ipython
>
> There should be shims for keep old commands and imports working during the
> transition, so let us know if you find places where that?s not the case.
>
> For more info, see the announcement
> <http://blog.jupyter.org/2015/08/12/first-release-of-jupyter/>, and the
> new jupyter docs <https://jupyter.readthedocs.org/en/latest/>.
>
> Thanks everyone!
>
> -The IPython and Jupyter Teams
> ?
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150812/75a80825/attachment.html>

From mattwcraig at gmail.com  Thu Aug 13 17:48:07 2015
From: mattwcraig at gmail.com (Matt Craig)
Date: Thu, 13 Aug 2015 16:48:07 -0500
Subject: [IPython-dev] How should requirements be listed in setup.py
	post-4.0?
Message-ID: <CAKxiDY=JgRpvVAfdOurVkwroagA-g5uKJU8F1j8phbQ2Dh1FHA@mail.gmail.com>

Hi,

I actually have two questions, both related to a widget-based
package I'm about release a new version of that is compatible with
both IPython 3 and IPython 4/jupyter.

1. What is the "best" thing to list as a requirement for the widget
bits...jupyter? ipywidgets?something else?

2. Anyone know a way to express in install_requires that either IPython v3
or jupyter/ipywidgets/whatever-the-answer-to1-is meets the install
requirements? I could fudge this by doing two separate releases, but would
rather not.

Thanks,
Matt Craig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150813/5e1af062/attachment.html>

From a.suchaneck at gmail.com  Fri Aug 14 13:32:30 2015
From: a.suchaneck at gmail.com (Anton Suchaneck)
Date: Fri, 14 Aug 2015 19:32:30 +0200
Subject: [IPython-dev] Debugging when communication of notebook with
 kernel lost?
In-Reply-To: <93AA912F-85E7-4D39-8278-F9F0E8BE9C9D@gmail.com>
References: <CA+ou_hXnRhxZfJNtZBgrDTu6V-UHTJ9axXb7rrBAYiygD89EcQ@mail.gmail.com>
	<93AA912F-85E7-4D39-8278-F9F0E8BE9C9D@gmail.com>
Message-ID: <CA+ou_hUumwrAU+M81+r6Sb1gQ2Ac8r=FOutgFQta=bsbjESP=Q@mail.gmail.com>

Hello,

I've now repeatedly had this issue in different systems, where the IPython
just stopped executing any further cells. Maybe I can try to tell what I
observed (since it's hard to reproduce not all points are clear to me):
* cells that usually run in milliseconds, just show [*] and nothing happens
anymore.
* there is no message on the IPython console or the Javascript console
* it happens on Firefox, Chrome, Windows, Linux (not just my own system). I
suppose all had an antivirus - one had only McAfee
* on the IPython console window the notebook continues to save itself
(while no cell is executed)
* I believe once it continued after a long pause, but I'm not sure
* kernel restarting from the menu usually does not fix it
* the only thing that seems to work is going to the IPython home screen,
shutdown the notebook and restart
* sometimes I see "KeyboardInterrupt" in the console at eventloops.py: line
30, process_stream_events()
* I have a vague impression it happens more often when I do plotting, but
that's just a guess

What would you suggest to find out what's going on?

Regards,
Anton


2015-06-27 15:31 GMT+02:00 Matthias Bussonnier <bussonniermatthias at gmail.com
>:

> Hi Anton,
>
>
> > On Jun 27, 2015, at 14:36, Anton Suchaneck <a.suchaneck at gmail.com>
> wrote:
> >
> > Hi!
> >
> > I occasionally have the issue that no more commands are being executed
> in the IPython notebook. The interface responds, but just leaves [*] cells
> without progressing or interrupting on request. Sometimes this even happens
> right after starting. I cannot find an error message. Often restarting the
> kernel works, but sometimes even that doesn't work right away.
>
>
> This look really like
> 1) the kernel is actually doing computation
> 2) the packets get intercepted/dropped in between fronted/server.
>
> I doubt this is 1, so it is most likely 2,
>
> unfortunately if it is 2, there will probably not be errors, are we are
> just waiting for things that will never come back.
>
> > How can I inspect what's going on (and maybe report that)?
>
> I would suggest starting the notebook with --debug, to get more verbose
> messages if any,
> and can you look at the Javascript console to see if there is anything ?
>
> > I work on a local Windows 7 computer with company firewall. Someone
> suggested it's due to Windows.
>
> Do you have sophos antivirus ?
>
> Cheers,
> --
> M
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150814/8015691e/attachment.html>

From p.f.moore at gmail.com  Mon Aug 17 08:30:17 2015
From: p.f.moore at gmail.com (Paul Moore)
Date: Mon, 17 Aug 2015 13:30:17 +0100
Subject: [IPython-dev] [ANN] Released IPython 4.0,
	completing The Big Split
In-Reply-To: <CAGXzvieZQBtV3YxRYVT0m1Tombq53_zHtYTJWYZs__wUyprSQQ@mail.gmail.com>
References: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
	<CAGXzvieZQBtV3YxRYVT0m1Tombq53_zHtYTJWYZs__wUyprSQQ@mail.gmail.com>
Message-ID: <CACac1F9WFYWCfy=C1thT+2Y42JDg5rKmm78L8K3n9tjAFLa+Ew@mail.gmail.com>

On 12 August 2015 at 05:30, David Powell
<DavidAnthonyPowell+python at gmail.com> wrote:
> Great news, and thanks for the hard work.
>
> I just used pip to upgrade my Windows conda based installation. One thing I
> noticed is that all of the jupyter scripts are python files without
> extensions, rather than windows .exe files. So "jupyter notebook" does not
> work from the command line. However "ipython notebook" works just fine, as
> there is a new ipython.exe in my scripts folder.
>
> Should the installer be creating this .exe file? If it's a bug, to which
> repo(s) should it be reported? It's not so clear after the big split.

(Disclaimer - I've not tried the new version yet) I would have
expected the jupyter command to be a setuptools entry point, so that
exe wrappers would be generated automatically.

That appears to be the case (from
https://github.com/jupyter/jupyter_core/blob/master/setup.py#L71) so
maybe there's something going wrong in your setup?
Paul


From jenshnielsen at gmail.com  Mon Aug 17 08:33:05 2015
From: jenshnielsen at gmail.com (Jens Nielsen)
Date: Mon, 17 Aug 2015 12:33:05 +0000
Subject: [IPython-dev] [ANN] Released IPython 4.0,
	completing The Big Split
In-Reply-To: <CACac1F9WFYWCfy=C1thT+2Y42JDg5rKmm78L8K3n9tjAFLa+Ew@mail.gmail.com>
References: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
	<CAGXzvieZQBtV3YxRYVT0m1Tombq53_zHtYTJWYZs__wUyprSQQ@mail.gmail.com>
	<CACac1F9WFYWCfy=C1thT+2Y42JDg5rKmm78L8K3n9tjAFLa+Ew@mail.gmail.com>
Message-ID: <CAM-Pw01WgbO6RmQPvnEboaugy9LneOS5iCeOfaYTqPxR5jc=YQ@mail.gmail.com>

That's because it's already been fixed :) The commit adding that is 4 days
old
https://github.com/jupyter/jupyter_core/commit/ffb50f81d7f5612135a25411b780465195da5021

Jens


man. 17. aug. 2015 kl. 13.30 skrev Paul Moore <p.f.moore at gmail.com>:

> On 12 August 2015 at 05:30, David Powell
> <DavidAnthonyPowell+python at gmail.com> wrote:
> > Great news, and thanks for the hard work.
> >
> > I just used pip to upgrade my Windows conda based installation. One
> thing I
> > noticed is that all of the jupyter scripts are python files without
> > extensions, rather than windows .exe files. So "jupyter notebook" does
> not
> > work from the command line. However "ipython notebook" works just fine,
> as
> > there is a new ipython.exe in my scripts folder.
> >
> > Should the installer be creating this .exe file? If it's a bug, to which
> > repo(s) should it be reported? It's not so clear after the big split.
>
> (Disclaimer - I've not tried the new version yet) I would have
> expected the jupyter command to be a setuptools entry point, so that
> exe wrappers would be generated automatically.
>
> That appears to be the case (from
> https://github.com/jupyter/jupyter_core/blob/master/setup.py#L71) so
> maybe there's something going wrong in your setup?
> Paul
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150817/b5fd8f45/attachment.html>

From cappy2112 at gmail.com  Mon Aug 17 12:57:48 2015
From: cappy2112 at gmail.com (Tony Cappellini)
Date: Mon, 17 Aug 2015 09:57:48 -0700
Subject: [IPython-dev] Session not found: session - seen again on Windows 7,
	64-Bit
Message-ID: <CABRXM4=b0pfW8zyw00Tvg9cw5-Gp0sW1sfWQ-js2NdbUiN08oQ@mail.gmail.com>

Hello,

I've been seeing problems on a Windows 7, 64-Bit system with the iPython
notebooks.
I'm currently running Python 2.7.10, and iPython 3.2.0

More than once, when I've tried closing a notebook, the notebook did not
close.
Today this issue occurred again, and I've found an error message in the
iP{ython console window.

[I 09:47:00.128 NotebookApp] Kernel started:
175ccc7a-8093-4a2a-8aeb-942a5b72990d
[I 09:47:20.006 NotebookApp] Kernel shutdown:
440f2ba4-bd02-4141-983d-afd8c2b0ff4b
[W 09:47:28.135 NotebookApp] Session not found:
session_id=u'9bd405d3-7123-4447-9e9d-3bfa64282646'
[W 09:47:28.135 NotebookApp] 404 DELETE
/api/sessions/9bd405d3-7123-4447-9e9d-3bfa64282646 (::1) 0.00ms referer=
http://localhost:8888/notebooks/August%2012%20-%20Daily%20Notebook-Installed%20BOOST%20package%20o
n%20Centos%20dev%20system%2C%20using%20yum.ipynb

I tend to use lengthy descriptions in the title field of my notebooks,
based on whatever activities
I was working on at the time I created the notebook. I am wondering if the
lengthy notebook name is causing a problem.I would expect that the file
name length limits be driven by the OS on each particular platform.

I realize the iPython 4.0 was just released, but I tend not to jump on the
bleeding edge, until they have been out there for a while.

Shall I file a bug report?

Thanks

Tony
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150817/6544f86e/attachment.html>

From p.f.moore at gmail.com  Mon Aug 17 15:39:28 2015
From: p.f.moore at gmail.com (Paul Moore)
Date: Mon, 17 Aug 2015 20:39:28 +0100
Subject: [IPython-dev] [ANN] Released IPython 4.0,
	completing The Big Split
In-Reply-To: <CAM-Pw01WgbO6RmQPvnEboaugy9LneOS5iCeOfaYTqPxR5jc=YQ@mail.gmail.com>
References: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
	<CAGXzvieZQBtV3YxRYVT0m1Tombq53_zHtYTJWYZs__wUyprSQQ@mail.gmail.com>
	<CACac1F9WFYWCfy=C1thT+2Y42JDg5rKmm78L8K3n9tjAFLa+Ew@mail.gmail.com>
	<CAM-Pw01WgbO6RmQPvnEboaugy9LneOS5iCeOfaYTqPxR5jc=YQ@mail.gmail.com>
Message-ID: <CACac1F8ZwBp2c3FfgqZyYhmCeOuw1f46rry58dPZmFCKxRPA8A@mail.gmail.com>

On 17 August 2015 at 13:33, Jens Nielsen <jenshnielsen at gmail.com> wrote:
> That's because it's already been fixed :) The commit adding that is 4 days
> old
> https://github.com/jupyter/jupyter_core/commit/ffb50f81d7f5612135a25411b780465195da5021

My bad for going on holiday and then answering old emails :-)

Thanks,
Paul


From wes.turner at gmail.com  Mon Aug 17 17:51:37 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Mon, 17 Aug 2015 16:51:37 -0500
Subject: [IPython-dev] [ANN] Released IPython 4.0,
	completing The Big Split
In-Reply-To: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
References: <CAHNn8BWJRw_nx5XW1OovAThT5GTbSuR4MBDAkWP8-MjTjxYnrg@mail.gmail.com>
Message-ID: <CACfEFw-F4-ncVdCiQjrCEUhLasvo4nT+PF2N7seZG12sWsk16Q@mail.gmail.com>

Thanks!

- [ ] DOC: how to rebuild docker containers with the same mapped e.g ~/ or
~/nb path (and/or remote storage extensions already-installed)
- [ ] DOC: update and/or include docker install instructions

https://github.com/ipython/ipython/wiki/Install:-Docker#links-to-jupyter-hub

We?ve just released IPython 4.0, completing the transition of
language-agnostic pieces of IPython to their new homes on Jupyter. This is
primarily a reorganization of the code into the new repos, with a few bugs
fixed along the way.

To install the new jupyter everything, including IPython, the notebook,
qtconsole, etc.:

pip install --upgrade jupyter

and start the notebook with:

jupyter notebook

or get just the IPython terminal, without any of the Jupyter fun:

pip install --upgrade ipython

There should be shims for keep old commands and imports working during the
transition, so let us know if you find places where that?s not the case.

For more info, see the announcement
<http://blog.jupyter.org/2015/08/12/first-release-of-jupyter/>, and
the new jupyter
docs <https://jupyter.readthedocs.org/en/latest/>.

Thanks everyone!

-The IPython and Jupyter Teams
?

_______________________________________________
IPython-dev mailing list
IPython-dev at scipy.org
http://mail.scipy.org/mailman/listinfo/ipython-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150817/3a9bd099/attachment.html>

From jon.freder at gmail.com  Mon Aug 17 21:19:28 2015
From: jon.freder at gmail.com (Jonathan Frederic)
Date: Mon, 17 Aug 2015 18:19:28 -0700
Subject: [IPython-dev] How should requirements be listed in setup.py
	post-4.0?
In-Reply-To: <CAKxiDY=JgRpvVAfdOurVkwroagA-g5uKJU8F1j8phbQ2Dh1FHA@mail.gmail.com>
References: <CAKxiDY=JgRpvVAfdOurVkwroagA-g5uKJU8F1j8phbQ2Dh1FHA@mail.gmail.com>
Message-ID: <CAAoBLw10ShnuqJ1Df=pJbGnoBRPGppt8XK+CQyUVa2Q1BVxc=w@mail.gmail.com>

Hi Matt,

Question 1 is a little hard to answer, seeing that you're targeting both
3.0 and 4.0.  For 3.0 ipython/ipython, for 4.0 ipython/ipywidgets (which
itself depends on ipython/ipython)

Question 2, I do not know the answer to.

Cheers,
Jon

On Thu, Aug 13, 2015 at 2:48 PM, Matt Craig <mattwcraig at gmail.com> wrote:

> Hi,
>
> I actually have two questions, both related to a widget-based
> package I'm about release a new version of that is compatible with
> both IPython 3 and IPython 4/jupyter.
>
> 1. What is the "best" thing to list as a requirement for the widget
> bits...jupyter? ipywidgets?something else?
>
> 2. Anyone know a way to express in install_requires that either IPython v3
> or jupyter/ipywidgets/whatever-the-answer-to1-is meets the install
> requirements? I could fudge this by doing two separate releases, but would
> rather not.
>
> Thanks,
> Matt Craig
>
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150817/3be7c711/attachment.html>

From madtom1999 at googlemail.com  Tue Aug 18 06:36:09 2015
From: madtom1999 at googlemail.com (Tom)
Date: Tue, 18 Aug 2015 11:36:09 +0100
Subject: [IPython-dev] Using turtle  and print with Notebook
Message-ID: <55D30A99.7090701@googlemail.com>

I'd like to use Notebook with a serious of lessons I'm writing to use 
turtle to take beginners from simple shapes to a full blown gui and all 
the bits in between.
Most of it works well in Notebook but if events on the turtle screen run 
print statements then they 'queue up until one runs a python action in 
one of the code cell when all the queued prints show up beneath.
Is there any way of getting these to appear as they arrive?
Tom



From erik.m.bray at gmail.com  Tue Aug 18 14:03:12 2015
From: erik.m.bray at gmail.com (Erik Bray)
Date: Tue, 18 Aug 2015 14:03:12 -0400
Subject: [IPython-dev] How should requirements be listed in setup.py
	post-4.0?
In-Reply-To: <CAKxiDY=JgRpvVAfdOurVkwroagA-g5uKJU8F1j8phbQ2Dh1FHA@mail.gmail.com>
References: <CAKxiDY=JgRpvVAfdOurVkwroagA-g5uKJU8F1j8phbQ2Dh1FHA@mail.gmail.com>
Message-ID: <CAOTD34b+wVzRe7tX1W5+_wPBX=QAGjJy=Mrqmxp1vRLNucWZPw@mail.gmail.com>

On Thu, Aug 13, 2015 at 5:48 PM, Matt Craig <mattwcraig at gmail.com> wrote:
> Hi,
>
> I actually have two questions, both related to a widget-based package I'm
> about release a new version of that is compatible with both IPython 3 and
> IPython 4/jupyter.
>
> 1. What is the "best" thing to list as a requirement for the widget
> bits...jupyter? ipywidgets?something else?
>
> 2. Anyone know a way to express in install_requires that either IPython v3
> or jupyter/ipywidgets/whatever-the-answer-to1-is meets the install
> requirements? I could fudge this by doing two separate releases, but would
> rather not.

This might be a good question for distutils-sig.  But I don't think
there's currently a great way currently to specify a requirement that
may be satisfied. PEP-426 provides for this, but that isn't currently
implemented in any practically usable way.

One kind of hacky approach would be to specify each dependency as an
extras_require [2].  Something like:

extras_require = {
    'IPython': ['IPython>=3.0'],
    'ipywidgets': ['ipywidgets'],
    'jupyter': ['juptyter']
}

This is sort of an abuse of the extras system since is actually a
*required* dependency (to have at least one).  But you could put
something in your setup.py that tries importing each one of them in
turn (or better, use pkg_resources.get_distribution or the like), and
if none of the required dependencies are found raise an exception
stating that at least one of IPython, ipywidgets, or jupyter must be
installed with the package, and you can select which one you want
using the syntax:

    $ pip install "MyPackage[jupyter]"

for example.

So this can be read sort of as "MyPackage with Jupyter support".  This
is similar to, but perhaps slightly nicer than having multiple
distributions for each possible dependency.

Erik

[1] https://www.python.org/dev/peps/pep-0426/#id40
[2] https://pythonhosted.org/setuptools/setuptools.html#declaring-extras-optional-features-with-their-own-dependencies


From robert.dodier at gmail.com  Tue Aug 18 20:37:47 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Tue, 18 Aug 2015 17:37:47 -0700
Subject: [IPython-dev] how to implement syntax highlighting for a new
	language
In-Reply-To: <CAOvn4qgF-qY+zsw0gv+EmKO57gALsE=JbbB=RjD1qVFgdCLzgg@mail.gmail.com>
References: <CAAsY_sQyn6j9Bv6-UcGf3kpBO80yFOJBc+2OcNK8EgJvaJhBfA@mail.gmail.com>
	<CAOvn4qgF-qY+zsw0gv+EmKO57gALsE=JbbB=RjD1qVFgdCLzgg@mail.gmail.com>
Message-ID: <CAAsY_sQf6JHn=fpKcUg33znoyuGN7i12WNgNhq=FG+hHBSNAMQ@mail.gmail.com>

Hi Thomas, hi everyone,

> Codemirror is the JS text editor we use for notebook cells. Many
> kernels can just specify one of the modes codemirror already
> provides: https://codemirror.net/mode/
>
> If there isn't a suitable mode already in codemirror (it doesn't look
> like there's one for Maxima), you'll need to provide one.

I have createdd a mode for Maxima by modifying one of the existing modes,
but it doesn't seem to be activated -- stuff is shown without highlighting.
I'm expecting that stuff I type into an input cell will be colorized --
that's the way it is supposed to work, right?

How is IPython supposed to activate or trigger CodeMirror?
Perhaps I'm missing something very obvious. I tried copying the
CodeMirror-related stuff from Fishbowl (the project from which I
forked Maxima-Fishbowl) but nothing happens. What I'm not
understanding is, what is the magic in IPython which causes
CodeMirror to be applied to stuff. Thanks in advance for any
light you can shed on this question.

best,

Robert Dodier


From bussonniermatthias at gmail.com  Wed Aug 19 05:11:59 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Wed, 19 Aug 2015 11:11:59 +0200
Subject: [IPython-dev] how to implement syntax highlighting for a new
	language
In-Reply-To: <CAAsY_sQf6JHn=fpKcUg33znoyuGN7i12WNgNhq=FG+hHBSNAMQ@mail.gmail.com>
References: <CAAsY_sQyn6j9Bv6-UcGf3kpBO80yFOJBc+2OcNK8EgJvaJhBfA@mail.gmail.com>
	<CAOvn4qgF-qY+zsw0gv+EmKO57gALsE=JbbB=RjD1qVFgdCLzgg@mail.gmail.com>
	<CAAsY_sQf6JHn=fpKcUg33znoyuGN7i12WNgNhq=FG+hHBSNAMQ@mail.gmail.com>
Message-ID: <CANJQusXr=G4opFhd0Sb9tavazkkqMjs7hGKwdC_HH1hhX0_8dA@mail.gmail.com>

Hi Robert,

Once your highlighting mode in in the right place in the codemirror
directory (notebook/components/codemirror/mode/modename/modefile.js
IIRC) you need your kernel-info-reply to give the name of the
codemirror mode is one of the field. Most of the time the language
name is enough, codemirror has a mapping Language<->mode.

Otherwise you can explicitely set the mode.
See http://jupyter-client.readthedocs.org/en/latest/messaging.html#kernel-info

If you can point us to where your mode is, and where your kernel is we
can help you to figure that out.
-- 
m


On Wed, Aug 19, 2015 at 2:37 AM, Robert Dodier <robert.dodier at gmail.com> wrote:
> Hi Thomas, hi everyone,
>
>> Codemirror is the JS text editor we use for notebook cells. Many
>> kernels can just specify one of the modes codemirror already
>> provides: https://codemirror.net/mode/
>>
>> If there isn't a suitable mode already in codemirror (it doesn't look
>> like there's one for Maxima), you'll need to provide one.
>
> I have createdd a mode for Maxima by modifying one of the existing modes,
> but it doesn't seem to be activated -- stuff is shown without highlighting.
> I'm expecting that stuff I type into an input cell will be colorized --
> that's the way it is supposed to work, right?
>
> How is IPython supposed to activate or trigger CodeMirror?
> Perhaps I'm missing something very obvious. I tried copying the
> CodeMirror-related stuff from Fishbowl (the project from which I
> forked Maxima-Fishbowl) but nothing happens. What I'm not
> understanding is, what is the magic in IPython which causes
> CodeMirror to be applied to stuff. Thanks in advance for any
> light you can shed on this question.
>
> best,
>
> Robert Dodier
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev


From ellisonbg at gmail.com  Wed Aug 19 13:11:33 2015
From: ellisonbg at gmail.com (Brian Granger)
Date: Wed, 19 Aug 2015 10:11:33 -0700
Subject: [IPython-dev] [jupyter] nbgrader version 0.2.0 released
In-Reply-To: <24614D05-5B93-4B96-85F2-9338D0B7DAF4@gmail.com>
References: <24614D05-5B93-4B96-85F2-9338D0B7DAF4@gmail.com>
Message-ID: <CAH4pYpSmFSD7GW5zMNz6DUUhx=QBdm2eGeJMiwb=Cxw7iSw8bA@mail.gmail.com>

This is great! Congrats on the release!

On Wed, Aug 19, 2015 at 9:13 AM, Jess Hamrick
<jessica.b.hamrick at gmail.com> wrote:
> Hi all,
>
> Version 0.2.0 of nbgrader primarily adds support for version 4.0 of the Jupyter notebook and associated project after The Big Split. The full list of major changes are:
>
>         ? Jupyter notebook 4.0 support
>         ? Make it possible to run the formgrader inside a Docker container
>         ? Make course_id a requirement in the transfer apps (list, release, fetch, submit, collect)
>         ? Add a new assignment list extension which allows students to list, fetch, validate, and submit assignments from the notebook dashboard interface
>         ? Auto-resize text boxes when giving feedback in the formgrader
>         ? Deprecate the BasicConfig and NbGraderConfig classes in favor of a NbGrader class
>
> Thanks to the following contributors who submitted PRs or reported issues that were merged/closed for the 0.2.0 release:
>
>         ? alope107
>         ? Carreau
>         ? ellisonbg
>         ? jhamrick
>         ? svurens
>
> Cheers,
> Jess
>
> --
> You received this message because you are subscribed to the Google Groups "Project Jupyter" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to jupyter+unsubscribe at googlegroups.com.
> To post to this group, send email to jupyter at googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/jupyter/24614D05-5B93-4B96-85F2-9338D0B7DAF4%40gmail.com.
> For more options, visit https://groups.google.com/d/optout.



-- 
Brian E. Granger
Associate Professor of Physics and Data Science
Cal Poly State University, San Luis Obispo
@ellisonbg on Twitter and GitHub
bgranger at calpoly.edu and ellisonbg at gmail.com


From robert.dodier at gmail.com  Wed Aug 19 19:30:35 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Wed, 19 Aug 2015 16:30:35 -0700
Subject: [IPython-dev] how to implement syntax highlighting for a new
	language
In-Reply-To: <CANJQusXr=G4opFhd0Sb9tavazkkqMjs7hGKwdC_HH1hhX0_8dA@mail.gmail.com>
References: <CAAsY_sQyn6j9Bv6-UcGf3kpBO80yFOJBc+2OcNK8EgJvaJhBfA@mail.gmail.com>
	<CAOvn4qgF-qY+zsw0gv+EmKO57gALsE=JbbB=RjD1qVFgdCLzgg@mail.gmail.com>
	<CAAsY_sQf6JHn=fpKcUg33znoyuGN7i12WNgNhq=FG+hHBSNAMQ@mail.gmail.com>
	<CANJQusXr=G4opFhd0Sb9tavazkkqMjs7hGKwdC_HH1hhX0_8dA@mail.gmail.com>
Message-ID: <CAAsY_sS3c8AUA9v2dz+qGCC+ZgrdkkgJDdWw1Zz7h7LyRPVG0g@mail.gmail.com>

On Wed, Aug 19, 2015 at 2:11 AM, Matthias Bussonnier
<bussonniermatthias at gmail.com> wrote:

> Once your highlighting mode in in the right place in the codemirror
> directory (notebook/components/codemirror/mode/modename/modefile.js
> IIRC)

I can't seem to get this to work for me. I'm getting a 404 for
maxima.js no matter where I put it, e.g.:

[W 16:16:10.332 NotebookApp] 404 GET
/static/components/codemirror/mode/maxima/maxima.js?v=20150819161553
(127.0.0.1) 1.66ms
referer=http://localhost:8889/notebooks/Untitled13.ipynb?kernel_name=maxima

I have tried putting maxima.js in the system-wide installation directory
for CodeMirror (and putting Maxima on the list in codemirror/mode/meta.js
for good measure), and also putting maxima.js into
 ~/.ipython/profile_default/static/components/codemirror/mode/maxima/.

Is there a way to query the notebook server to find out what
its document root is? With an older version of IPython (namely 2.x)
that I was working with, it appeared that the document root was
 ~/.ipython/profile_something/, but that doesn't seem to be the
case anymore. From where is the notebook server trying to
load static/something?

How is the profile selected, by the way? and does it matter that
it is in a directory named ~/.ipython instead of ~/.jupyter?
I find ~/.jupyter is empty except for a file named "migrated".

> you need your kernel-info-reply to give the name of the
> codemirror mode is one of the field. Most of the time the language
> name is enough, codemirror has a mapping Language<->mode.

I think this part is working as expected -- the Maxima kernel puts
"maxima" as the language-info-name and then Jupyter tries to
load maxima.js as I hoped it would.

Thank you very much for your help, I appreciate it.

best,

Robert Dodier


From fperez.net at gmail.com  Wed Aug 19 19:35:42 2015
From: fperez.net at gmail.com (Fernando Perez)
Date: Wed, 19 Aug 2015 16:35:42 -0700
Subject: [IPython-dev] [jupyter] nbgrader version 0.2.0 released
In-Reply-To: <24614D05-5B93-4B96-85F2-9338D0B7DAF4@gmail.com>
References: <24614D05-5B93-4B96-85F2-9338D0B7DAF4@gmail.com>
Message-ID: <CAHAreOqrDd83roe4Wp=4--3MsZj+oZ7fRTV7XzNHC7S7rBDHXA@mail.gmail.com>

Congrats, Jess, delighted to see this progress (and that you eventually
beat Travis CI after yesterday's battles ;)

Cheers,

f

On Wed, Aug 19, 2015 at 9:13 AM, Jess Hamrick <jessica.b.hamrick at gmail.com>
wrote:

> Hi all,
>
> Version 0.2.0 of nbgrader primarily adds support for version 4.0 of the
> Jupyter notebook and associated project after The Big Split. The full list
> of major changes are:
>
>         ? Jupyter notebook 4.0 support
>         ? Make it possible to run the formgrader inside a Docker container
>         ? Make course_id a requirement in the transfer apps (list,
> release, fetch, submit, collect)
>         ? Add a new assignment list extension which allows students to
> list, fetch, validate, and submit assignments from the notebook dashboard
> interface
>         ? Auto-resize text boxes when giving feedback in the formgrader
>         ? Deprecate the BasicConfig and NbGraderConfig classes in favor of
> a NbGrader class
>
> Thanks to the following contributors who submitted PRs or reported issues
> that were merged/closed for the 0.2.0 release:
>
>         ? alope107
>         ? Carreau
>         ? ellisonbg
>         ? jhamrick
>         ? svurens
>
> Cheers,
> Jess
>
> --
> You received this message because you are subscribed to the Google Groups
> "Project Jupyter" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to jupyter+unsubscribe at googlegroups.com.
> To post to this group, send email to jupyter at googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/jupyter/24614D05-5B93-4B96-85F2-9338D0B7DAF4%40gmail.com
> .
> For more options, visit https://groups.google.com/d/optout.
>



-- 
Fernando Perez (@fperez_org; http://fperez.org)
fperez.net-at-gmail: mailing lists only (I ignore this when swamped!)
fernando.perez-at-berkeley: contact me here for any direct mail
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150819/9e68ce4a/attachment.html>

From wes.turner at gmail.com  Wed Aug 19 20:15:50 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Wed, 19 Aug 2015 19:15:50 -0500
Subject: [IPython-dev] [jupyter] nbgrader version 0.2.0 released
In-Reply-To: <CAH4pYpSmFSD7GW5zMNz6DUUhx=QBdm2eGeJMiwb=Cxw7iSw8bA@mail.gmail.com>
References: <24614D05-5B93-4B96-85F2-9338D0B7DAF4@gmail.com>
	<CAH4pYpSmFSD7GW5zMNz6DUUhx=QBdm2eGeJMiwb=Cxw7iSw8bA@mail.gmail.com>
Message-ID: <CACfEFw8WRtvnGaWGpkvwd3hY6dCcnnZmTZ9dbLAFua2etKgTeQ@mail.gmail.com>

On Aug 19, 2015 12:12 PM, "Brian Granger" <ellisonbg at gmail.com> wrote:
>
> This is great! Congrats on the release!
>
> On Wed, Aug 19, 2015 at 9:13 AM, Jess Hamrick
> <jessica.b.hamrick at gmail.com> wrote:
> > Hi all,
> >
> > Version 0.2.0 of nbgrader primarily adds support for version 4.0 of the
Jupyter notebook and associated project after The Big Split. The full list
of major changes are:
> >
> >         ? Jupyter notebook 4.0 support
> >         ? Make it possible to run the formgrader inside a Docker
container

Thanks!

> >         ? Make course_id a requirement in the transfer apps (list,
release, fetch, submit, collect)

How feasible would course_id as a URI be?

> >         ? Add a new assignment list extension which allows students to
list, fetch, validate, and submit assignments from the notebook dashboard
interface
> >         ? Auto-resize text boxes when giving feedback in the formgrader
> >         ? Deprecate the BasicConfig and NbGraderConfig classes in favor
of a NbGrader class
> >
> > Thanks to the following contributors who submitted PRs or reported
issues that were merged/closed for the 0.2.0 release:
> >
> >         ? alope107
> >         ? Carreau
> >         ? ellisonbg
> >         ? jhamrick
> >         ? svurens
> >
> > Cheers,
> > Jess
> >
> > --
> > You received this message because you are subscribed to the Google
Groups "Project Jupyter" group.
> > To unsubscribe from this group and stop receiving emails from it, send
an email to jupyter+unsubscribe at googlegroups.com.
> > To post to this group, send email to jupyter at googlegroups.com.
> > To view this discussion on the web visit
https://groups.google.com/d/msgid/jupyter/24614D05-5B93-4B96-85F2-9338D0B7DAF4%40gmail.com
.
> > For more options, visit https://groups.google.com/d/optout.
>
>
>
> --
> Brian E. Granger
> Associate Professor of Physics and Data Science
> Cal Poly State University, San Luis Obispo
> @ellisonbg on Twitter and GitHub
> bgranger at calpoly.edu and ellisonbg at gmail.com
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150819/9d1eaaca/attachment.html>

From fred.jendrzejewski at gmail.com  Thu Aug 20 02:25:11 2015
From: fred.jendrzejewski at gmail.com (Fred Jendrzejewski)
Date: Thu, 20 Aug 2015 08:25:11 +0200
Subject: [IPython-dev] Question about the stability of the api
Message-ID: <75B6B38C-E429-42E1-875B-A82401A9EE82@gmail.com>

Hello everyone,

I started working with a third party package (pymc3) in ipython4 I get some errors that seem concern changes in the api of ipython:

ShimWarning: The `IPython.html` package has been deprecated. You should 
import from `notebook` instead. `IPython.html.widgets` has moved to
`ipywidgets`. "`IPython.html.widgets` has moved to `ipywidgets`.", ShimWarning)
./lib/python3.4/site-packages/IPython/utils/traitlets.py:5: UserWarning: IPython.utils.traitlets 
has moved to a top-level traitlets package.
warn("IPython.utils.traitlets has moved to a top-level traitlets package."
Before we wanted to start the debugging procedure, we wanted to ask how stable this part of the ipython API is. In other words, is it worth to find a clean fix or will this change within a short time frame anyway again ?

I hope that you can help me.

Cheers,

Fred
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150820/70ec8f5b/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3595 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150820/70ec8f5b/attachment.bin>

From bussonniermatthias at gmail.com  Thu Aug 20 02:53:31 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Thu, 20 Aug 2015 08:53:31 +0200
Subject: [IPython-dev] Question about the stability of the api
In-Reply-To: <75B6B38C-E429-42E1-875B-A82401A9EE82@gmail.com>
References: <75B6B38C-E429-42E1-875B-A82401A9EE82@gmail.com>
Message-ID: <EAF574A2-A94C-4D48-B729-072E07203362@gmail.com>

Hi Fred, 


> On Aug 20, 2015, at 08:25, Fred Jendrzejewski <fred.jendrzejewski at gmail.com> wrote:
> 
> Hello everyone,
> 
> I started working with a third party package (pymc3) in ipython4 I get some errors that seem concern changes in the api of ipython:
> 
> ShimWarning: The `IPython.html` package has been deprecated. You should 
> import from `notebook` instead. `IPython.html.widgets` has moved to
> `ipywidgets`. "`IPython.html.widgets` has moved to `ipywidgets`.", ShimWarning)
> ./lib/python3.4/site-packages/IPython/utils/traitlets.py:5: UserWarning: IPython.utils.traitlets 
> has moved to a top-level traitlets package.
> warn("IPython.utils.traitlets has moved to a top-level traitlets package."
> Before we wanted to start the debugging procedure, we wanted to ask how stable this part of the ipython API is. In other words, is it worth to find a clean fix or will this change within a short time frame anyway again ?
> 

These are just warnings

You can just replace 

import IPython.html.widgets 

by 

try: 
   import ipywidgets
except ImportErrror:
  import IPython.html.widgets.

(same with traitlets)

Or actually do a conditional import depending on the IPython version. 
  
The stability depends on each package, that know have their own version numbers, so the stability will depends on each specific package. 
They will most likely not move again before a few years. 
-- 
M
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150820/02a57a46/attachment.html>

From fred.jendrzejewski at gmail.com  Thu Aug 20 02:56:31 2015
From: fred.jendrzejewski at gmail.com (Fred Jendrzejewski)
Date: Thu, 20 Aug 2015 08:56:31 +0200
Subject: [IPython-dev] Question about the stability of the api
In-Reply-To: <EAF574A2-A94C-4D48-B729-072E07203362@gmail.com>
References: <75B6B38C-E429-42E1-875B-A82401A9EE82@gmail.com>
	<EAF574A2-A94C-4D48-B729-072E07203362@gmail.com>
Message-ID: <435D3B2C-29DC-463A-80E9-7A4CC4F94DAD@gmail.com>

Thanks this helps a lot.

?Fred

> Am 20.08.2015 um 08:53 schrieb Matthias Bussonnier <bussonniermatthias at gmail.com>:
> 
> Hi Fred, 
> 
> 
>> On Aug 20, 2015, at 08:25, Fred Jendrzejewski <fred.jendrzejewski at gmail.com <mailto:fred.jendrzejewski at gmail.com>> wrote:
>> 
>> Hello everyone,
>> 
>> I started working with a third party package (pymc3) in ipython4 I get some errors that seem concern changes in the api of ipython:
>> 
>> ShimWarning: The `IPython.html` package has been deprecated. You should 
>> import from `notebook` instead. `IPython.html.widgets` has moved to
>> `ipywidgets`. "`IPython.html.widgets` has moved to `ipywidgets`.", ShimWarning)
>> ./lib/python3.4/site-packages/IPython/utils/traitlets.py:5: UserWarning: IPython.utils.traitlets 
>> has moved to a top-level traitlets package.
>> warn("IPython.utils.traitlets has moved to a top-level traitlets package."
>> Before we wanted to start the debugging procedure, we wanted to ask how stable this part of the ipython API is. In other words, is it worth to find a clean fix or will this change within a short time frame anyway again ?
>> 
> 
> These are just warnings
> 
> You can just replace 
> 
> import IPython.html.widgets 
> 
> by 
> 
> try: 
>    import ipywidgets
> except ImportErrror:
>   import IPython.html.widgets.
> 
> (same with traitlets)
> 
> Or actually do a conditional import depending on the IPython version. 
>   
> The stability depends on each package, that know have their own version numbers, so the stability will depends on each specific package. 
> They will most likely not move again before a few years. 
> -- 
> M
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150820/3dbd5710/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3595 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150820/3dbd5710/attachment.bin>

From wes.turner at gmail.com  Thu Aug 20 03:47:30 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Thu, 20 Aug 2015 02:47:30 -0500
Subject: [IPython-dev] Question about the stability of the api
In-Reply-To: <EAF574A2-A94C-4D48-B729-072E07203362@gmail.com>
References: <75B6B38C-E429-42E1-875B-A82401A9EE82@gmail.com>
	<EAF574A2-A94C-4D48-B729-072E07203362@gmail.com>
Message-ID: <CACfEFw-8ya_OQSFK1xFKJSExxOB_NN6tBsYQKzyM=diohacjXg@mail.gmail.com>

On Aug 20, 2015 1:53 AM, "Matthias Bussonnier" <bussonniermatthias at gmail.com>
wrote:
>
> Hi Fred,
>
>
>> On Aug 20, 2015, at 08:25, Fred Jendrzejewski <
fred.jendrzejewski at gmail.com> wrote:
>>
>> Hello everyone,
>>
>> I started working with a third party package (pymc3) in ipython4 I get
some errors that seem concern changes in the api of ipython:
>>
>> ShimWarning: The `IPython.html` package has been deprecated. You should
>> import from `notebook` instead. `IPython.html.widgets` has moved to
>> `ipywidgets`. "`IPython.html.widgets` has moved to `ipywidgets`.",
ShimWarning)
>> ./lib/python3.4/site-packages/IPython/utils/traitlets.py:5: UserWarning:
IPython.utils.traitlets
>> has moved to a top-level traitlets package.
>> warn("IPython.utils.traitlets has moved to a top-level traitlets
package."
>>
>> Before we wanted to start the debugging procedure, we wanted to ask how
stable this part of the ipython API is. In other words, is it worth to find
a clean fix or will this change within a short time frame anyway again ?
>
>
> These are just warnings
>
> You can just replace
>
> import IPython.html.widgets
>
> by
>
> try:
>    import ipywidgets
> except ImportErrror:
>   import IPython.html.widgets.
>
> (same with traitlets)
>
> Or actually do a conditional import depending on the IPython version.
>
> The stability depends on each package, that know have their own version
numbers, so the stability will depends on each specific package.
> They will most likely not move again before a few years.

http://sphinx-doc.org/markup/para.html

http://docs.ckan.org/en/latest/contributing/documentation.html#versionadded-and-versionchanged

.. versionchanged:
..

IPython widgets

| docs:
http://nbviewer.ipython.org/github/ipython/ipython/blob/master/examples/Interactive%20Widgets/Index.ipynb
| docs:
http://nbviewer.ipython.org/github/ipython/ipython/blob/master/examples/Interactive%20Widgets/Custom%20Widget%20-%20Hello%20World.ipynb
| docs: https://ipython.org/ipython-doc/dev/api/
| docs:
https://ipython.org/ipython-doc/dev/api/generated/IPython.html.widgets.widget.html
| src:
https://github.com/ipython/ipython/tree/master/examples/Interactive%20Widgets
| src: https://github.com/ipython/ipython/tree/3.x/IPython/html/widgets

IPython/ipywidgets

| docs:
http://nbviewer.ipython.org/github/ipython/ipywidgets/blob/master/examples/Index.ipynb
| docs:
http://nbviewer.ipython.org/github/ipython/ipywidgets/tree/master/examples/
| src: https://github.com/ipython/ipywidgets/tree/master/examples
| src: https://github.com/ipython/ipywidgets
| src: https://github.com/ipython/ipywidgets/tree/master/ipywidgets/widgets

Also,

* [ ] view source links in the IPython API docs (conf.py)?
   * [ ] view source links to (-> github at revtag?)

> --
> M
>
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150820/87b28192/attachment.html>

From maidos93 at laposte.net  Fri Aug 21 10:10:02 2015
From: maidos93 at laposte.net (thwiouz)
Date: Fri, 21 Aug 2015 07:10:02 -0700 (PDT)
Subject: [IPython-dev] Install matlab kernel on windows
Message-ID: <1440166202428-5169864.post@n6.nabble.com>

Hi,

I'm using Matlab 2014 on a W8 machine and updated my conda env to use
jupyter notebook. I tried to install the matlab_kernel and everything
worked.
But when trying the "ipython qtconsole --kernel=matlab_kernel" command,
Jupyter says me that it couldn't find the matlab_kernel.

Am I missing something?

Thanks in advance,



--
View this message in context: http://python.6.x6.nabble.com/Install-matlab-kernel-on-windows-tp5169864.html
Sent from the IPython - Development mailing list archive at Nabble.com.


From bussonniermatthias at gmail.com  Fri Aug 21 10:24:17 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Fri, 21 Aug 2015 16:24:17 +0200
Subject: [IPython-dev] Install matlab kernel on windows
In-Reply-To: <1440166202428-5169864.post@n6.nabble.com>
References: <1440166202428-5169864.post@n6.nabble.com>
Message-ID: <2C904584-D05D-4719-8EEF-7B09CB5A1A9B@gmail.com>


> On Aug 21, 2015, at 16:10, thwiouz <maidos93 at laposte.net> wrote:
> 
> Hi,
> 
> I'm using Matlab 2014 on a W8 machine and updated my conda env to use
> jupyter notebook. I tried to install the matlab_kernel and everything
> worked.
> But when trying the "ipython qtconsole --kernel=matlab_kernel" command,
> Jupyter says me that it couldn't find the matlab_kernel.

can you try 
$ ipython kernelspec list ?

If matlab does not appear in the list, then there is most-likely an registering step you missed or that failed during setup. 
You can also try to open an issue on matlab_kernel bug tracker, they might now better how matlab_kernel register with IPython. 

-- 
M



From maidos93 at laposte.net  Fri Aug 21 10:29:54 2015
From: maidos93 at laposte.net (thwiouz)
Date: Fri, 21 Aug 2015 07:29:54 -0700 (PDT)
Subject: [IPython-dev] Install matlab kernel on windows
In-Reply-To: <2C904584-D05D-4719-8EEF-7B09CB5A1A9B@gmail.com>
References: <1440166202428-5169864.post@n6.nabble.com>
	<2C904584-D05D-4719-8EEF-7B09CB5A1A9B@gmail.com>
Message-ID: <1440167394780-5169867.post@n6.nabble.com>

Ok thank you. Actually you're right, it doesn't appear.

I'll issued on the github project.

Thanks a lot;



--
View this message in context: http://python.6.x6.nabble.com/Install-matlab-kernel-on-windows-tp5169864p5169867.html
Sent from the IPython - Development mailing list archive at Nabble.com.


From maidos93 at laposte.net  Fri Aug 21 12:01:48 2015
From: maidos93 at laposte.net (thwiouz)
Date: Fri, 21 Aug 2015 09:01:48 -0700 (PDT)
Subject: [IPython-dev] Install matlab kernel on windows
In-Reply-To: <1440167394780-5169867.post@n6.nabble.com>
References: <1440166202428-5169864.post@n6.nabble.com>
	<2C904584-D05D-4719-8EEF-7B09CB5A1A9B@gmail.com>
	<1440167394780-5169867.post@n6.nabble.com>
Message-ID: <1440172908679-5169882.post@n6.nabble.com>

Just wanted to close the question as I now installed it from Github sources
and works.
Carefull that the pip version does not seem to work.

Thanks Matthias



--
View this message in context: http://python.6.x6.nabble.com/Install-matlab-kernel-on-windows-tp5169864p5169882.html
Sent from the IPython - Development mailing list archive at Nabble.com.


From robert.dodier at gmail.com  Fri Aug 21 18:35:09 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Fri, 21 Aug 2015 15:35:09 -0700
Subject: [IPython-dev] how to implement syntax highlighting for a new
	language
In-Reply-To: <CAAsY_sS3c8AUA9v2dz+qGCC+ZgrdkkgJDdWw1Zz7h7LyRPVG0g@mail.gmail.com>
References: <CAAsY_sQyn6j9Bv6-UcGf3kpBO80yFOJBc+2OcNK8EgJvaJhBfA@mail.gmail.com>
	<CAOvn4qgF-qY+zsw0gv+EmKO57gALsE=JbbB=RjD1qVFgdCLzgg@mail.gmail.com>
	<CAAsY_sQf6JHn=fpKcUg33znoyuGN7i12WNgNhq=FG+hHBSNAMQ@mail.gmail.com>
	<CANJQusXr=G4opFhd0Sb9tavazkkqMjs7hGKwdC_HH1hhX0_8dA@mail.gmail.com>
	<CAAsY_sS3c8AUA9v2dz+qGCC+ZgrdkkgJDdWw1Zz7h7LyRPVG0g@mail.gmail.com>
Message-ID: <CAAsY_sSE-QR1z=7eU0qAhk4ur_pmuAsNtEt3o_0phtZFbMnWTw@mail.gmail.com>

On Wed, Aug 19, 2015 at 4:30 PM, Robert Dodier <robert.dodier at gmail.com> wrote:

> I can't seem to get this to work for me. I'm getting a 404 for
> maxima.js no matter where I put it, e.g.:

I did get it to work, finally. By running jupyter with --debug I found
where JS was being loaded from, and put my CodeMirror mode
there. Voila! It works! Incidentally on my system, the location is:

/usr/local/lib/python3.4/dist-packages/notebook/static/components/codemirror/mode

Of course it might be different on other systems.

I was trying to install the CodeMirror mode somewhere in ~/.ipython
instead of a system directory, but I couldn't get that to work.
Maybe someone else has more luck with that.

Anyway, now that it's working I'm happy!

best

Robert Dodier


From ssilvester at continuum.io  Sat Aug 22 11:54:36 2015
From: ssilvester at continuum.io (Steven Silvester)
Date: Sat, 22 Aug 2015 10:54:36 -0500
Subject: [IPython-dev] Install matlab kernel on windows
In-Reply-To: <mailman.3.1440176401.19744.ipython-dev@scipy.org>
References: <mailman.3.1440176401.19744.ipython-dev@scipy.org>
Message-ID: <9977257E-1EA6-44A4-92E5-358896D6459C@continuum.io>

Thanks for the heads up thwiouz, I released new versions of Matlab Kernel and Octave Kernel that should fix the pip install problem.

Regards,

Steve



> On Aug 21, 2015, at 12:00 PM, ipython-dev-request at scipy.org wrote:
> 
> Send IPython-dev mailing list submissions to
> 	ipython-dev at scipy.org
> 
> To subscribe or unsubscribe via the World Wide Web, visit
> 	http://mail.scipy.org/mailman/listinfo/ipython-dev
> or, via email, send a message with subject or body 'help' to
> 	ipython-dev-request at scipy.org
> 
> You can reach the person managing the list at
> 	ipython-dev-owner at scipy.org
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of IPython-dev digest..."
> 
> 
> Today's Topics:
> 
>   1. Install matlab kernel on windows (thwiouz)
>   2. Re: Install matlab kernel on windows (Matthias Bussonnier)
>   3. Re: Install matlab kernel on windows (thwiouz)
>   4. Re: Install matlab kernel on windows (thwiouz)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Fri, 21 Aug 2015 07:10:02 -0700 (PDT)
> From: thwiouz <maidos93 at laposte.net>
> Subject: [IPython-dev] Install matlab kernel on windows
> To: ipython-dev at scipy.org
> Message-ID: <1440166202428-5169864.post at n6.nabble.com>
> Content-Type: text/plain; charset=us-ascii
> 
> Hi,
> 
> I'm using Matlab 2014 on a W8 machine and updated my conda env to use
> jupyter notebook. I tried to install the matlab_kernel and everything
> worked.
> But when trying the "ipython qtconsole --kernel=matlab_kernel" command,
> Jupyter says me that it couldn't find the matlab_kernel.
> 
> Am I missing something?
> 
> Thanks in advance,
> 
> 
> 
> --
> View this message in context: http://python.6.x6.nabble.com/Install-matlab-kernel-on-windows-tp5169864.html
> Sent from the IPython - Development mailing list archive at Nabble.com.
> 
> 
> ------------------------------
> 
> Message: 2
> Date: Fri, 21 Aug 2015 16:24:17 +0200
> From: Matthias Bussonnier <bussonniermatthias at gmail.com>
> Subject: Re: [IPython-dev] Install matlab kernel on windows
> To: IPython developers list <ipython-dev at scipy.org>
> Message-ID: <2C904584-D05D-4719-8EEF-7B09CB5A1A9B at gmail.com>
> Content-Type: text/plain; charset=us-ascii
> 
> 
>> On Aug 21, 2015, at 16:10, thwiouz <maidos93 at laposte.net> wrote:
>> 
>> Hi,
>> 
>> I'm using Matlab 2014 on a W8 machine and updated my conda env to use
>> jupyter notebook. I tried to install the matlab_kernel and everything
>> worked.
>> But when trying the "ipython qtconsole --kernel=matlab_kernel" command,
>> Jupyter says me that it couldn't find the matlab_kernel.
> 
> can you try 
> $ ipython kernelspec list ?
> 
> If matlab does not appear in the list, then there is most-likely an registering step you missed or that failed during setup. 
> You can also try to open an issue on matlab_kernel bug tracker, they might now better how matlab_kernel register with IPython. 
> 
> -- 
> M
> 
> 
> 
> ------------------------------
> 
> Message: 3
> Date: Fri, 21 Aug 2015 07:29:54 -0700 (PDT)
> From: thwiouz <maidos93 at laposte.net>
> Subject: Re: [IPython-dev] Install matlab kernel on windows
> To: ipython-dev at scipy.org
> Message-ID: <1440167394780-5169867.post at n6.nabble.com>
> Content-Type: text/plain; charset=us-ascii
> 
> Ok thank you. Actually you're right, it doesn't appear.
> 
> I'll issued on the github project.
> 
> Thanks a lot;
> 
> 
> 
> --
> View this message in context: http://python.6.x6.nabble.com/Install-matlab-kernel-on-windows-tp5169864p5169867.html
> Sent from the IPython - Development mailing list archive at Nabble.com.
> 
> 
> ------------------------------
> 
> Message: 4
> Date: Fri, 21 Aug 2015 09:01:48 -0700 (PDT)
> From: thwiouz <maidos93 at laposte.net>
> Subject: Re: [IPython-dev] Install matlab kernel on windows
> To: ipython-dev at scipy.org
> Message-ID: <1440172908679-5169882.post at n6.nabble.com>
> Content-Type: text/plain; charset=us-ascii
> 
> Just wanted to close the question as I now installed it from Github sources
> and works.
> Carefull that the pip version does not seem to work.
> 
> Thanks Matthias
> 
> 
> 
> --
> View this message in context: http://python.6.x6.nabble.com/Install-matlab-kernel-on-windows-tp5169864p5169882.html
> Sent from the IPython - Development mailing list archive at Nabble.com.
> 
> 
> ------------------------------
> 
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
> 
> 
> End of IPython-dev Digest, Vol 139, Issue 19
> ********************************************

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150822/971ae81f/attachment.html>

From bussonniermatthias at gmail.com  Sun Aug 23 04:04:06 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Sun, 23 Aug 2015 10:04:06 +0200
Subject: [IPython-dev] how to implement syntax highlighting for a new
	language
In-Reply-To: <CAAsY_sSE-QR1z=7eU0qAhk4ur_pmuAsNtEt3o_0phtZFbMnWTw@mail.gmail.com>
References: <CAAsY_sQyn6j9Bv6-UcGf3kpBO80yFOJBc+2OcNK8EgJvaJhBfA@mail.gmail.com>
	<CAOvn4qgF-qY+zsw0gv+EmKO57gALsE=JbbB=RjD1qVFgdCLzgg@mail.gmail.com>
	<CAAsY_sQf6JHn=fpKcUg33znoyuGN7i12WNgNhq=FG+hHBSNAMQ@mail.gmail.com>
	<CANJQusXr=G4opFhd0Sb9tavazkkqMjs7hGKwdC_HH1hhX0_8dA@mail.gmail.com>
	<CAAsY_sS3c8AUA9v2dz+qGCC+ZgrdkkgJDdWw1Zz7h7LyRPVG0g@mail.gmail.com>
	<CAAsY_sSE-QR1z=7eU0qAhk4ur_pmuAsNtEt3o_0phtZFbMnWTw@mail.gmail.com>
Message-ID: <CANJQusV7G=EUc6D3c=c4Ygmh_XYsWLsw4jjpLX-7xTFgJwO6iQ@mail.gmail.com>

Ok, great, sorry a bit overwhelmed the last few days.
Still on my todo list to do.

Once your mode is in codemirror itself, if shoudl automatically
install in /usr/local/lib/python3.4/dist-packages/notebook/static/components/codemirror/mode
when codemirror is installed.
-- 
M

On Sat, Aug 22, 2015 at 12:35 AM, Robert Dodier <robert.dodier at gmail.com> wrote:
> On Wed, Aug 19, 2015 at 4:30 PM, Robert Dodier <robert.dodier at gmail.com> wrote:
>
>> I can't seem to get this to work for me. I'm getting a 404 for
>> maxima.js no matter where I put it, e.g.:
>
> I did get it to work, finally. By running jupyter with --debug I found
> where JS was being loaded from, and put my CodeMirror mode
> there. Voila! It works! Incidentally on my system, the location is:
>
> /usr/local/lib/python3.4/dist-packages/notebook/static/components/codemirror/mode
>
> Of course it might be different on other systems.
>
> I was trying to install the CodeMirror mode somewhere in ~/.ipython
> instead of a system directory, but I couldn't get that to work.
> Maybe someone else has more luck with that.
>
> Anyway, now that it's working I'm happy!
>
> best
>
> Robert Dodier
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev


From akim at lrde.epita.fr  Mon Aug 24 11:39:18 2015
From: akim at lrde.epita.fr (Akim Demaille)
Date: Mon, 24 Aug 2015 17:39:18 +0200
Subject: [IPython-dev] [IPython-User] Compressed notebooks
In-Reply-To: <88771671-2D79-433F-9140-3CB0D20D2130@gmail.com>
References: <242CFF6C-F06D-4F08-912E-4234E8C3D84E@lrde.epita.fr>
	<88771671-2D79-433F-9140-3CB0D20D2130@gmail.com>
Message-ID: <DACB44DC-80D5-47EC-AE38-E5B046394FC1@lrde.epita.fr>


> Le 24 ao?t 2015 ? 15:48, Matthias Bussonnier <bussonniermatthias at gmail.com> a ?crit :
> 
> Hi Akim, 

Hi!

> We don?t want to compress notebook as otherwise they become not human-fixable anymore, 
> and even worse to version control.

I agree with that.  I would never commit anything compressed.  Man pages and Info pages are of course never committed compressed.  However, man and info readers grok compressed files.  The case of installed documentation (which is readonly btw, it's only about _reading_ compressed, not writing), is really different.  But then again, I'm not asking for this feature, just checking whether you considered this use case.

> I believe the Julia Taylor might have already done some work for packaging notebooks, 
> you might want to contact him, or coordinate, if you haven?t already. 

No, I haven't.  I'll first check to see if there already information about that somewhere, thanks!

From bussonniermatthias at gmail.com  Mon Aug 24 12:25:34 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Mon, 24 Aug 2015 18:25:34 +0200
Subject: [IPython-dev] [IPython-User] Compressed notebooks
In-Reply-To: <DACB44DC-80D5-47EC-AE38-E5B046394FC1@lrde.epita.fr>
References: <242CFF6C-F06D-4F08-912E-4234E8C3D84E@lrde.epita.fr>
	<88771671-2D79-433F-9140-3CB0D20D2130@gmail.com>
	<DACB44DC-80D5-47EC-AE38-E5B046394FC1@lrde.epita.fr>
Message-ID: <CANJQusWikz-bfw54r3WEcDWtKvQouHVKawy+m77m8DHw=Mhy=A@mail.gmail.com>

Hey,


>> I believe the Julia Taylor might have already done some work for packaging notebooks,
>> you might want to contact him, or coordinate, if you haven?t already.
>
> No, I haven't.  I'll first check to see if there already information about that somewhere, thanks!

Typo from my part: Julian Taylor. Her is his PPA.
https://launchpad.net/~jtaylor/+archive/ubuntu/ipython


> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev


From wes.turner at gmail.com  Mon Aug 24 14:57:32 2015
From: wes.turner at gmail.com (Wes Turner)
Date: Mon, 24 Aug 2015 13:57:32 -0500
Subject: [IPython-dev] ENH: git-wip work in progress branches
Message-ID: <CACfEFw_nq9gdoTO+xgZTUTnQfY16XAqABgNSDbiM5yFzYROV4Q@mail.gmail.com>

"git-wip [1] is a script that will manage Work In Progress (or WIP)
branches. [2] WIP branches are mostly throw away but identify points of
development between commits. The intent is to tie this script into your
editor so that each time you save your file, the git-wip script captures
that state in git. git-wip also helps you return back to a previous state
of development." [1]

Where and how could we integrate git-wip with Jupyter Notebook?

* [ ] [re-]implement git-wip in Python (w/w/o bindings)?
* [ ] PY: extend/plugin on_autosave() / on_save()
* [ ] JS: add commit message dialogue
* [ ] JS: add log + revert to graph node dialogue [2]
* [ ] PY: package as an extension with jupyter-pip [3]


[1] https://github.com/bartman/git-wip

[2] https://github.com/bartman/git-wip#wip-branches

[3] https://github.com/jdfreder/jupyter-pip

[*] https://github.com/joeyh/etckeeper has similar goals.


---

frontend:
*
https://github.com/jupyter/notebook/blob/master/notebook/tests/notebook/save.js
*
https://github.com/jupyter/notebook/blob/master/notebook/static/notebook/js/savewidget.js
*
https://github.com/jupyter/notebook/blob/master/notebook/static/notebook/js/notebook.js#L2071
"Notebook.prototype.save_notebook"


backend:
*
https://github.com/jupyter/notebook/blob/master/notebook/services/contents/filemanager.py#L82
"post_save_hook"


---

Is there something like git-wip branches for Jupyter already?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150824/d94a1944/attachment.html>

From takowl at gmail.com  Mon Aug 24 15:17:06 2015
From: takowl at gmail.com (Thomas Kluyver)
Date: Mon, 24 Aug 2015 12:17:06 -0700
Subject: [IPython-dev] ENH: git-wip work in progress branches
In-Reply-To: <CACfEFw_nq9gdoTO+xgZTUTnQfY16XAqABgNSDbiM5yFzYROV4Q@mail.gmail.com>
References: <CACfEFw_nq9gdoTO+xgZTUTnQfY16XAqABgNSDbiM5yFzYROV4Q@mail.gmail.com>
Message-ID: <CAOvn4qjv761oa2Lq+TqMo3ttZMVJcMjjQ-ewx_6mg3mJ1yUeCw@mail.gmail.com>

On 24 August 2015 at 11:57, Wes Turner <wes.turner at gmail.com> wrote:

> * [ ] [re-]implement git-wip in Python (w/w/o bindings)?
>
I don't think this is necessary. It's currently in bash, which is fine
unless people want to use it on Windows (and that's not a problem specific
to Jupyter).


> * [ ] PY: extend/plugin on_autosave() / on_save()


We already have save hooks - I just opened a PR documenting them. I don't
think they know about autosave vs manual save , though.

Thomas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150824/3800dbc2/attachment.html>

From mark.voorhies at ucsf.edu  Mon Aug 24 15:20:21 2015
From: mark.voorhies at ucsf.edu (Mark Voorhies)
Date: Mon, 24 Aug 2015 12:20:21 -0700
Subject: [IPython-dev] ENH: git-wip work in progress branches
In-Reply-To: <CACfEFw_nq9gdoTO+xgZTUTnQfY16XAqABgNSDbiM5yFzYROV4Q@mail.gmail.com>
References: <CACfEFw_nq9gdoTO+xgZTUTnQfY16XAqABgNSDbiM5yFzYROV4Q@mail.gmail.com>
Message-ID: <55DB6E75.8000504@ucsf.edu>

On 08/24/2015 11:57 AM, Wes Turner wrote:
> "git-wip [1] is a script that will manage Work In Progress (or WIP)
> branches. [2] WIP branches are mostly throw away but identify points of
> development between commits. The intent is to tie this script into your
> editor so that each time you save your file, the git-wip script captures
> that state in git. git-wip also helps you return back to a previous state
> of development." [1]
>
> Where and how could we integrate git-wip with Jupyter Notebook?
>
> * [ ] [re-]implement git-wip in Python (w/w/o bindings)?

Might be useful to have a look at flashbake 
(https://github.com/commandline/flashbake/wiki), which is already in 
python.  Cory Doctorow
has been using this for quite a while to auto-vc his writing.

--Mark


From robert.dodier at gmail.com  Mon Aug 24 19:59:09 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Mon, 24 Aug 2015 16:59:09 -0700
Subject: [IPython-dev] how to display messages received by console or
	notebook front end
Message-ID: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>

Hi, I'm trying to implement input_request/input_reply in a kernel for
Maxima. Is there a way to get the console or notebook front end to
display the messages it receives from the kernel? The
--Session.debug=True option seems to cause it to show only the
messages sent from the front end to the kernel, and not in the other
direction.

By the way, how is the console or notebook supposed to respond to input_request?

Thanks for any info,

Robert Dodier


From bussonniermatthias at gmail.com  Tue Aug 25 03:54:17 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Tue, 25 Aug 2015 09:54:17 +0200
Subject: [IPython-dev] how to display messages received by console or
	notebook front end
In-Reply-To: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
References: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
Message-ID: <0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>


> On Aug 25, 2015, at 01:59, Robert Dodier <robert.dodier at gmail.com> wrote:
> 
> Hi, I'm trying to implement input_request/input_reply in a kernel for
> Maxima. Is there a way to get the console or notebook front end to
> display the messages it receives from the kernel? The
> --Session.debug=True option seems to cause it to show only the
> messages sent from the front end to the kernel, and not in the other
> direction.

We just updated jupyter_client master to do that.
https://github.com/jupyter/jupyter_client/pull/64 <https://github.com/jupyter/jupyter_client/pull/64>

> 
> By the way, how is the console or notebook supposed to respond to input_request?

The notebook append an in put field which is submitted when the user press enter.
you can try on try.jupyter.org with the following in a python cell:

name = input('name:?)

If the fronted does not support input request it should raise a not implemented from our side in Python at least IIRC. 

Does that reply to your question ?
-- 
M

> 
> Thanks for any info,
> 
> Robert Dodier
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150825/7a09e1f7/attachment.html>

From robert.dodier at gmail.com  Tue Aug 25 21:03:46 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Tue, 25 Aug 2015 18:03:46 -0700
Subject: [IPython-dev] how to display messages received by console or
 notebook front end
In-Reply-To: <0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>
References: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
	<0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>
Message-ID: <CAAsY_sTu-rng56nCnHwVSHnVNABOf=D5XPz1YEvtwKA_gJmTkQ@mail.gmail.com>

Thanks a lot, Matthias. I applied the patch and now I see messages
received by the front end (as well as messages sent).

The problem I have been banging my head against is that I am trying to
send messages on the stdin channel from the kernel to the front end
and the messages don't appear to be sent. This is very mysterious to
me. I am working with a Lisp wrapper around libzmq and I can trace
calls to the point where the wrapper is supposed to call zmq_send but
nothing gets sent. I am watching the stdin port via tcpdump and there
is no traffic there when I try to send something from my Maxima
kernel, but there is traffic when I use the Python kernel and enter
input("foo") as you suggested.

I tried strace to see if I could see socket writes but I don't see
them (perhaps I'm just not finding them in all the strace output).

The libzmq wrapper seems to be working OK since traffic on other
channels is sent and received successfully.

I know this isn't much to go on, but do you have any advice about how
I can debug this problem? Does it sound familiar, like any problem you
have encountered in the past?

Thanks for your continued assistance, I really appreciate it.

best,

Robert Dodier


From bussonniermatthias at gmail.com  Wed Aug 26 04:43:28 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Wed, 26 Aug 2015 10:43:28 +0200
Subject: [IPython-dev] how to display messages received by console or
	notebook front end
In-Reply-To: <CAAsY_sTu-rng56nCnHwVSHnVNABOf=D5XPz1YEvtwKA_gJmTkQ@mail.gmail.com>
References: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
	<0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>
	<CAAsY_sTu-rng56nCnHwVSHnVNABOf=D5XPz1YEvtwKA_gJmTkQ@mail.gmail.com>
Message-ID: <05DA2514-EA38-4490-9736-5B29F6AA614F@gmail.com>


> On Aug 26, 2015, at 03:03, Robert Dodier <robert.dodier at gmail.com> wrote:
> 
> Thanks a lot, Matthias. I applied the patch and now I see messages
> received by the front end (as well as messages sent).

Great !

> 
> The problem I have been banging my head against is that I am trying to
> send messages on the stdin channel from the kernel to the front end
> and the messages don't appear to be sent. This is very mysterious to
> me. I am working with a Lisp wrapper around libzmq and I can trace
> calls to the point where the wrapper is supposed to call zmq_send but
> nothing gets sent. I am watching the stdin port via tcpdump and there
> is no traffic there when I try to send something from my Maxima
> kernel, but there is traffic when I use the Python kernel and enter
> input("foo") as you suggested.


IPython?s input is not standard python input ! We replace the built-in with that:
https://github.com/ipython/ipykernel/blob/29f27ec7e4a21c19332364fede6b4276aeddc89b/ipykernel/kernelbase.py#L637-L652 <https://github.com/ipython/ipykernel/blob/29f27ec7e4a21c19332364fede6b4276aeddc89b/ipykernel/kernelbase.py#L637-L652>

(rest of the logic is in the file if it helps). 
Would any other request be blocked ?
Do you maybe need to reply to another in progress request on another channel before
sending things on stdin ?

I think one of the tricky bit with input_request is it is reversed from the other channel, 
like the bind/connect might be reversed as the kernel initiate a input request. 


> I tried strace to see if I could see socket writes but I don't see
> them (perhaps I'm just not finding them in all the strace output).
> 
> The libzmq wrapper seems to be working OK since traffic on other
> channels is sent and received successfully.
> 
> I know this isn't much to go on, but do you have any advice about how
> I can debug this problem? Does it sound familiar, like any problem you
> have encountered in the past?

I?ll keep thinking about it. 
If you have no answer let?s try to ping the jupyter google group where other(non python) kernel authors might have an answer. 

> Thanks for your continued assistance, I really appreciate it.

No problem, thanks for writing a maxima kernel !
And if any docs is unclear or need addition, please send us PRs !
-- 
M

> 
> best,
> 
> Robert Dodier
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150826/1c156c70/attachment.html>

From robert.dodier at gmail.com  Wed Aug 26 20:49:57 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Wed, 26 Aug 2015 17:49:57 -0700
Subject: [IPython-dev] how to display messages received by console or
 notebook front end
In-Reply-To: <05DA2514-EA38-4490-9736-5B29F6AA614F@gmail.com>
References: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
	<0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>
	<CAAsY_sTu-rng56nCnHwVSHnVNABOf=D5XPz1YEvtwKA_gJmTkQ@mail.gmail.com>
	<05DA2514-EA38-4490-9736-5B29F6AA614F@gmail.com>
Message-ID: <CAAsY_sQ68wpOunyu4JH1YJMYhT7jCxdU5+bxzr8ciRW-fxPGMw@mail.gmail.com>

Hi Matthias, hi everybody,

well, I got the input_request to work -- it turns out the crucial bit
which I missed is that the stdin socket must be a DEALER (I had
declared it as a ROUTER). Perhaps that datum will be helpful to
someone else.

With that change, TCP/IP traffic on the stdin socket does appear and
the front end does receive the input_request. The notebook opens a
text input prompt as expected and the kernel receives the input_reply.
O happy day!

best,

Robert Dodier


From bussonniermatthias at gmail.com  Thu Aug 27 03:45:47 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Thu, 27 Aug 2015 09:45:47 +0200
Subject: [IPython-dev] how to display messages received by console or
 notebook front end
In-Reply-To: <CAAsY_sQ68wpOunyu4JH1YJMYhT7jCxdU5+bxzr8ciRW-fxPGMw@mail.gmail.com>
References: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
	<0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>
	<CAAsY_sTu-rng56nCnHwVSHnVNABOf=D5XPz1YEvtwKA_gJmTkQ@mail.gmail.com>
	<05DA2514-EA38-4490-9736-5B29F6AA614F@gmail.com>
	<CAAsY_sQ68wpOunyu4JH1YJMYhT7jCxdU5+bxzr8ciRW-fxPGMw@mail.gmail.com>
Message-ID: <CANJQusUvQ-W64Vzoqn_D39bX19SyqZbWvjGRkjb17HB4=f3CFQ@mail.gmail.com>

Ok, Great !

if you think the docs should insist more on that somewhere, tell us !
(or send PRs !)



On Thu, Aug 27, 2015 at 2:49 AM, Robert Dodier <robert.dodier at gmail.com> wrote:
> Hi Matthias, hi everybody,
>
> well, I got the input_request to work -- it turns out the crucial bit
> which I missed is that the stdin socket must be a DEALER (I had
> declared it as a ROUTER). Perhaps that datum will be helpful to
> someone else.
>
> With that change, TCP/IP traffic on the stdin socket does appear and
> the front end does receive the input_request. The notebook opens a
> text input prompt as expected and the kernel receives the input_reply.
> O happy day!
>
> best,
>
> Robert Dodier
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev


From robert.dodier at gmail.com  Thu Aug 27 04:00:31 2015
From: robert.dodier at gmail.com (Robert Dodier)
Date: Thu, 27 Aug 2015 01:00:31 -0700
Subject: [IPython-dev] how to display messages received by console or
 notebook front end
In-Reply-To: <CANJQusUvQ-W64Vzoqn_D39bX19SyqZbWvjGRkjb17HB4=f3CFQ@mail.gmail.com>
References: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
	<0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>
	<CAAsY_sTu-rng56nCnHwVSHnVNABOf=D5XPz1YEvtwKA_gJmTkQ@mail.gmail.com>
	<05DA2514-EA38-4490-9736-5B29F6AA614F@gmail.com>
	<CAAsY_sQ68wpOunyu4JH1YJMYhT7jCxdU5+bxzr8ciRW-fxPGMw@mail.gmail.com>
	<CANJQusUvQ-W64Vzoqn_D39bX19SyqZbWvjGRkjb17HB4=f3CFQ@mail.gmail.com>
Message-ID: <CAAsY_sQ5QMq3RtVwzze6-XdMYC7fAih=o4QjmZ+h31rqu2r4EA@mail.gmail.com>

On Thu, Aug 27, 2015 at 12:45 AM, Matthias Bussonnier
<bussonniermatthias at gmail.com> wrote:

> if you think the docs should insist more on that somewhere, tell us !

Well, maybe there is something I've misunderstood. But the
documentation seems to pretty clearly say that the kernel's stdin
socket is a ROUTER.

E.g. http://jupyter-client.readthedocs.org/en/latest/messaging.html
says: "stdin: this ROUTER socket is connected to all frontends, and it
allows the kernel to request input from the active frontend when
raw_input() is called. The frontend that executed the code has a
DEALER socket that acts as a ?virtual keyboard? for the kernel ..."

But in order for it to work for Maxima at least, the kernel's stdin
needs to be a DEALER, otherwise the input_request message is not
delivered to the front end. (Just to be specific, the input_request
prompt is something like "Is k equal to -1?" when trying to solve,
say, integrate(x^k, x). This interaction is a hold-over from the
original design of Maxima, which assumed that there is always a user
sitting at the keyboard, to supply information to help the
computation.)

Thanks for your help,

Robert Dodier


From bussonniermatthias at gmail.com  Thu Aug 27 04:30:01 2015
From: bussonniermatthias at gmail.com (Matthias Bussonnier)
Date: Thu, 27 Aug 2015 10:30:01 +0200
Subject: [IPython-dev] how to display messages received by console or
 notebook front end
In-Reply-To: <CAAsY_sQ5QMq3RtVwzze6-XdMYC7fAih=o4QjmZ+h31rqu2r4EA@mail.gmail.com>
References: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
	<0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>
	<CAAsY_sTu-rng56nCnHwVSHnVNABOf=D5XPz1YEvtwKA_gJmTkQ@mail.gmail.com>
	<05DA2514-EA38-4490-9736-5B29F6AA614F@gmail.com>
	<CAAsY_sQ68wpOunyu4JH1YJMYhT7jCxdU5+bxzr8ciRW-fxPGMw@mail.gmail.com>
	<CANJQusUvQ-W64Vzoqn_D39bX19SyqZbWvjGRkjb17HB4=f3CFQ@mail.gmail.com>
	<CAAsY_sQ5QMq3RtVwzze6-XdMYC7fAih=o4QjmZ+h31rqu2r4EA@mail.gmail.com>
Message-ID: <CANJQusXCEmbmOOEK32U67G0R129DXCk1rV=hDQmMtyk4NkbpWQ@mail.gmail.com>

I'll defer on that to Min that probably can answer immediately while I
would have to die into that for hours.

In jupyter_client, stdin is definitively a DEALER, and on ipkernel a
ROUTER, then I don't know why you would have to reverse.

One drawback I can think of is that you might get issues with multiple
clients connected at once if you invert the two.
-- 
M




On Thu, Aug 27, 2015 at 10:00 AM, Robert Dodier <robert.dodier at gmail.com> wrote:
> On Thu, Aug 27, 2015 at 12:45 AM, Matthias Bussonnier
> <bussonniermatthias at gmail.com> wrote:
>
>> if you think the docs should insist more on that somewhere, tell us !
>
> Well, maybe there is something I've misunderstood. But the
> documentation seems to pretty clearly say that the kernel's stdin
> socket is a ROUTER.
>
> E.g. http://jupyter-client.readthedocs.org/en/latest/messaging.html
> says: "stdin: this ROUTER socket is connected to all frontends, and it
> allows the kernel to request input from the active frontend when
> raw_input() is called. The frontend that executed the code has a
> DEALER socket that acts as a ?virtual keyboard? for the kernel ..."
>
> But in order for it to work for Maxima at least, the kernel's stdin
> needs to be a DEALER, otherwise the input_request message is not
> delivered to the front end. (Just to be specific, the input_request
> prompt is something like "Is k equal to -1?" when trying to solve,
> say, integrate(x^k, x). This interaction is a hold-over from the
> original design of Maxima, which assumed that there is always a user
> sitting at the keyboard, to supply information to help the
> computation.)
>
> Thanks for your help,
>
> Robert Dodier
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev


From vasco+python at tenner.nl  Fri Aug 28 06:01:42 2015
From: vasco+python at tenner.nl (Vasco)
Date: Fri, 28 Aug 2015 12:01:42 +0200
Subject: [IPython-dev] executing cells before kernel is started
Message-ID: <55E03186.8050204@tenner.nl>

Dear all,
When I load a notebook, it takes some time before the kernel is started. 
When I try to execute cells in this time, they are not executed.

My workflow is often like this: I have all kinds of imports in the first 
cell, that I want to run. In the meantime, I scroll down to the part of 
the notebook I want to edit.

However, when I run the first cell to early, I findout halfway my 
notebook that it did not run.

It look likes that jupiter just drops all runned cells if there is no 
kernel. It would be very handy to queue all cells until the kernel is 
ready (at least at startup, before any cell has run).

Is there a way to obtain this behaviour?

Kind regards,
Vasco


From benjaminrk at gmail.com  Fri Aug 28 10:00:51 2015
From: benjaminrk at gmail.com (MinRK)
Date: Fri, 28 Aug 2015 07:00:51 -0700
Subject: [IPython-dev] how to display messages received by console or
 notebook front end
In-Reply-To: <CANJQusXCEmbmOOEK32U67G0R129DXCk1rV=hDQmMtyk4NkbpWQ@mail.gmail.com>
References: <CAAsY_sRG5BRzzxXb4Z1tfVEGEo5+RMKyPUTDMYY+0=+D8DL1Jw@mail.gmail.com>
	<0C3E6CAE-61BC-4ED7-96EC-69C1D64700AD@gmail.com>
	<CAAsY_sTu-rng56nCnHwVSHnVNABOf=D5XPz1YEvtwKA_gJmTkQ@mail.gmail.com>
	<05DA2514-EA38-4490-9736-5B29F6AA614F@gmail.com>
	<CAAsY_sQ68wpOunyu4JH1YJMYhT7jCxdU5+bxzr8ciRW-fxPGMw@mail.gmail.com>
	<CANJQusUvQ-W64Vzoqn_D39bX19SyqZbWvjGRkjb17HB4=f3CFQ@mail.gmail.com>
	<CAAsY_sQ5QMq3RtVwzze6-XdMYC7fAih=o4QjmZ+h31rqu2r4EA@mail.gmail.com>
	<CANJQusXCEmbmOOEK32U67G0R129DXCk1rV=hDQmMtyk4NkbpWQ@mail.gmail.com>
Message-ID: <CAHNn8BW+N9n4-Km9xynFkof=twVvtXi-rNeDAPn+fZeVY1v=hA@mail.gmail.com>

On Thu, Aug 27, 2015 at 1:30 AM, Matthias Bussonnier <
bussonniermatthias at gmail.com> wrote:

> I'll defer on that to Min that probably can answer immediately while I
> would have to die into that for hours.
>
> In jupyter_client, stdin is definitively a DEALER, and on ipkernel a
> ROUTER, then I don't know why you would have to reverse.
>
> One drawback I can think of is that you might get issues with multiple
> clients connected at once if you invert the two.
>

Yup, stdin should definitely be ROUTER on the kernel and DEALER on the
client. For this to work, it is important that the stdin DEALER on the
client be given the same IDENTITY as the shell DEALER on the same client.

-MinRK


> --
> M
>
>
>
>
> On Thu, Aug 27, 2015 at 10:00 AM, Robert Dodier <robert.dodier at gmail.com>
> wrote:
> > On Thu, Aug 27, 2015 at 12:45 AM, Matthias Bussonnier
> > <bussonniermatthias at gmail.com> wrote:
> >
> >> if you think the docs should insist more on that somewhere, tell us !
> >
> > Well, maybe there is something I've misunderstood. But the
> > documentation seems to pretty clearly say that the kernel's stdin
> > socket is a ROUTER.
> >
> > E.g. http://jupyter-client.readthedocs.org/en/latest/messaging.html
> > says: "stdin: this ROUTER socket is connected to all frontends, and it
> > allows the kernel to request input from the active frontend when
> > raw_input() is called. The frontend that executed the code has a
> > DEALER socket that acts as a ?virtual keyboard? for the kernel ..."
> >
> > But in order for it to work for Maxima at least, the kernel's stdin
> > needs to be a DEALER, otherwise the input_request message is not
> > delivered to the front end. (Just to be specific, the input_request
> > prompt is something like "Is k equal to -1?" when trying to solve,
> > say, integrate(x^k, x). This interaction is a hold-over from the
> > original design of Maxima, which assumed that there is always a user
> > sitting at the keyboard, to supply information to help the
> > computation.)
> >
> > Thanks for your help,
> >
> > Robert Dodier
> > _______________________________________________
> > IPython-dev mailing list
> > IPython-dev at scipy.org
> > http://mail.scipy.org/mailman/listinfo/ipython-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150828/110fe70e/attachment.html>

From benjaminrk at gmail.com  Fri Aug 28 10:27:12 2015
From: benjaminrk at gmail.com (MinRK)
Date: Fri, 28 Aug 2015 07:27:12 -0700
Subject: [IPython-dev] executing cells before kernel is started
In-Reply-To: <55E03186.8050204@tenner.nl>
References: <55E03186.8050204@tenner.nl>
Message-ID: <CAHNn8BUr4vLMFRG9ZwyC1iF-Ljn7zHkk+Lgy9EocnPc1gGyO0A@mail.gmail.com>

Yeah, the behavior of the frontend is not great when the kernel connection
hasn't been established. I think right now, it just refuses to execute
without comment. At the very least, it should make a note of the failure.

On Fri, Aug 28, 2015 at 3:01 AM, Vasco <vasco+python at tenner.nl> wrote:

> Dear all,
> When I load a notebook, it takes some time before the kernel is started.
> When I try to execute cells in this time, they are not executed.
>
> My workflow is often like this: I have all kinds of imports in the first
> cell, that I want to run. In the meantime, I scroll down to the part of
> the notebook I want to edit.
>
> However, when I run the first cell to early, I findout halfway my
> notebook that it did not run.
>
> It look likes that jupiter just drops all runned cells if there is no
> kernel. It would be very handy to queue all cells until the kernel is
> ready (at least at startup, before any cell has run).
>
> Is there a way to obtain this behaviour?
>
> Kind regards,
> Vasco
> _______________________________________________
> IPython-dev mailing list
> IPython-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ipython-dev/attachments/20150828/2488283c/attachment.html>