[Pythonmac-SIG] Python 2.3 vs 2.2, VTK, and Carbon

[Brian Lenihan]

Bob's right.  It does take forever to build, but VTK from CVS built and  
passed make test with no modifications using gcc 3.3.   Either there is  
something different about the build environment prior to Panther which  
causes the missing math.h reference, or I got lucky and answered all  
the CMake questions correctly.  I turned on "advanced questions" and  
answered the questions in what seemed to be the logical fashion.  I did  
a Release build, as I presume a debug build would fail without a  
corresponding debug Python framework.

I have Apple's and Fink's Python binaries renamed to avoid getting any  
conflicts and CMake seemed to make the correct choices even though it  
asked some problematic questions like where is my Python debug  
framework? even after I specified a Release build.  I used a framework  
build of Tcl/Tk 8.4 and only saw one bogus default CMake default answer  
referring to my Fink Tcl/Tk install.  I ignored it because it seemed  
harmless since CMake had previously correctly inferred I was using a  
framework build of Tcl/Tk

I can import vtk and a dir(vtk) looks OK.  A real test will have to  
wait for another day.   The August Dev tools released yesterday may  
also work.

On Sep 17, 2003, at 10:26 PM, Bob Ippolito wrote:

> Add:
> #include <math.h>
> before
> #include <Carbon/Carbon.h>
> ( or whatever the equivalent is.. wherever the error is coming from on  
> the VTK side )
>
> That should work.
>
> I tried compiling VTK once or twice, but it takes WAY too long to  
> compile on my 1ghz pbg4 (even with a gig of ram).. I've got a dual g5  
> on order, so maybe when that gets in (a week or two) I'll give it  
> another go.  Hopefully it'll be fast enough to tolerate multiple  
> compiles so I can fix the thing, then I'll add it to my PackageManager  
> repository to save everyone the trouble.
>
> -bob
>
> On Wednesday, Sep 17, 2003, at 23:21 America/New_York, W.T. Bridgman  
> wrote:
>
>> The verdict is that it happened again.  Same error.
>>
>> If this is while building the python wrapper, could it be a Tk issue?  
>>  One odd item about my  installation is that I used PackageManager to  
>> install the _tkinter.so file on one system, then copied the binary  
>> over to the other machine where I'm installing VTK (which doesn't go  
>> online).  However, I have a Tcl/Tk framework left over from when I  
>> tried out AlphaTk and CMake found that framework and used it.   
>> However, the version numbers seem to match (8.4).  Could that still  
>> create a conflict somewhere?
>>
>> On Wednesday, September 17, 2003, at 10:41  PM, W.T. Bridgman wrote:
>>
>>> I ran into exactly this error yesterday evening and am trying again  
>>> with yesterday evening's build at this very instant.  There were  
>>> some python wrapper files checked into the VTK repository.
>>>
>>> I'll try to keep you posted if I still have a problem but this build  
>>> is on my home system so I can only check it in the evening.   
>>> However, we've got a hurricane coming through sometime tomorrow  
>>> afternoon so don't know when I'll be back online.
>>>
>>> On Wednesday, August 6, 2003, at 10:09  PM, Yves Starreveld wrote:
>>>
>>>> One odd thing is that the importation of both Carbon/Carbon.h and  
>>>> Python.h from 2.3 leads to the following errors:
>>>>
>>>> /System/Library/Frameworks/CoreServices.framework/Frameworks/ 
>>>> CarbonCore.framework/Headers/fp.h: In
>>>>    function `long double scalbl(long double, long int)':
>>>> /System/Library/Frameworks/CoreServices.framework/Frameworks/ 
>>>> CarbonCore.framework/Headers/fp.h:1896: error: `
>>>>    scalb' undeclared (first use this function)
>>>> /System/Library/Frameworks/CoreServices.framework/Frameworks/ 
>>>> CarbonCore.framework/Headers/fp.h:1896: error: (Each
>>>>    undeclared identifier is reported only once for each function it  
>>>> appears
>>>>    in.)
>>>>
>>>>
>>>> Other files in vtk import Carbon.h without difficulty, so I am  
>>>> surmising a conflict. Has anyone run into anything similar?
>
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig
>