On 4/2/12 9:58 AM, Andrew Ackerman wrote:
> On 3/31/12 12:36 PM, Alexander Hansen wrote:
>> On 3/30/12 2:16 PM, Andrew Ackerman wrote:
>>> Sorry for the spam, but forgot to mention in my last message that I
>>> was able to update octave out of the box on another Intel Mac Pro,
>>> which in principle has the same OS setup, though it does have a
>>> different history, which may include some past variants on X11 (though
>>> now using official Apple version). So you may well not be able to
>>> reproduce the problem I had on the other Mac.
>>> -Andy
>>>
>>> On 3/30/12 5:12 PM, Andrew Ackerman wrote:
>>>> Thanks, Alexander -- your workaround worked.
>>>>
>>>> You ask about octave-x11. Answer is no, I was just updating the
>>>> vanilla flavor
>>>> of octave. Last time I looked at the list (could be some years ago),
>>>> there were
>>>> not so many octaves in fink. Quite a proliferation -- I don't know
>>>> what to make
>>>> of them all. Nothing jumps out at me from
>>>> http://pdb.finkproject.org/pdb/package.php/octave-x11 that
>>>> distinguishes it from
>>>> regular old octave.
>>>>
>>>> Thanks again,
>>>> Andy
>>>>
>>>> On 3/30/12 3:48 PM, Alexander Hansen wrote:
>>>>> On 3/30/12 12:35 PM, Andrew Ackerman wrote:
>>>>>> I've been using fink for many years (have stuck with 32-bit 
>>>>>> version),
>>>>>> and have always been able to get things to build before. Maybe I
>>>>>> should just be happy with my 3.0.3 version. I use the official Apple
>>>>>> X11, and my OS is completely up-to-date. I'm using a Mac Pro with 2
>>>>>> quad-core Intel Xeons (uname -a: Darwin gs611-altocu.giss.nasa.gov
>>>>>> 10.8.0 Darwin Kernel Version 10.8.0: Tue Jun 7 16:33:36 PDT 2011;
>>>>>> root:xnu-1504.15.3~1/RELEASE_I386 i386).
>>>>>>
>>>>>> I've followed the instructions for the failure (fink selfupdate and
>>>>>> MaxBuildJobs: 1), twice, with no progress.
>>>>>
>>>>> (in this case setting MaxBuildJobs doesn't do anything because I 
>>>>> had to
>>>>> force Octave to build with only one build job)
>>>>>>
>>>>>> It seems like the install is nearly complete, with the next to last
>>>>>> build commands looking good:
>>>>>>
>>>>>> ---
>>>>>> Installing .gdbinit from version at etc/gdbinit
>>>>>>
>>>>>> Octave successfully built. Now choose from the following:
>>>>>>
>>>>>> ./run-octave - to run in place to test before installing
>>>>>> make check - to run the tests
>>>>>> make install - to install (PREFIX=/sw)
>>>>>>
>>>>>> fink-package-precedence .
>>>>>> Scanning /\.Plo,\.Po$/ dependency files...
>>>>>> ---
>>>>>>
>>>>>> But then it gacks on the next build command:
>>>>>>
>>>>>> ---
>>>>>> Looking for incorrect headers in 1027 dependency files...
>>>>>> ./src/.deps/liboctinterp_la-txt-eng-ft.Plo uses
>>>>>> /usr/X11/include/fontconfig/fontconfig.h
>>>>>> Please fix build process to get consistent use of fink's headers.
>>>>>> ### execution of /tmp/fink.h7C1G failed, exit code 255
>>>>>> ---
>>>>>>
>>>>>> Anything not too hard I can do to fix this problem? I suppose I 
>>>>>> could
>>>>>> try reinstalling X11 a la
>>>>>> http://www.finkproject.org/faq/usage-packages.php?phpLang=en#special-x11-debug
>>>>>>  
>>>>>>
>>>>>>
>>>>>> but I'd rather not.
>>>>>>
>>>>>> Thanks,
>>>>>> Andy
>>>>> This is octave-x11, I assume? I'll see if I can reproduce the 
>>>>> issue on
>>>>> my own 10.6/32-bit setup.
>>>>>
>>>>> Reinstalling X11 isn't likely to change anything, so don't do 
>>>>> that. The
>>>>> failure is a complaint to the maintainer (me) that something in the
>>>>> build procedure is amiss according to our internal policies. :-) (I
>>>>> didn't test the X11 build too much this time around)
>>>>>
>>>>> A quick workaround is to edit
>>>>> /sw/fink/10.4/stable/main/finkinfo/sci/octave-3.6.1.info and comment
>>>>> out
>>>>> or remove the line that says " fink-package-precedence . " (under
>>>>> CompileScript). That should get you a functional Octave, but one that
>>>>> happens to link to X11's fontconfig library rather than Fink's.
>>>>>
>> Hi.  If you get some time and spare CPU, I've got a potential fix for
>> the problem which I'd appreciate your testing out:
>>
>> 1)  Remove /sw/fink/10.4/stable/main/finkinfo/sci/octave-3.6.1.info and
>> do a "fink selfupdate" to restore the file to what is in Fink right now
>>
>> -or-
>>
>> 1A) Restore the " fink-package-precedence . " line.
>>
>> 2) Edit octave-3.6.1.info again and add the following line in the block
>> above the CompileScript (maybe above the SetLDFLAGS line):
>>
>> SetCPPFLAGS: -I%p/lib/fontconfig2/include/fontconfig
>>
>> That should explicitly tell the build to look in our fontconfig2 header
>> directory first.
>>
>> 3)  Use "fink rebuild octave".  This won't disrupt your ability to use
>> what you've already built unless the build actually succeeds ;-) .
>
> Hi Alexander,
>
> Back to work now and I just tried your suggestion,
>
> Looking for incorrect headers in 1027 dependency files...
>          ./src/.deps/liboctinterp_la-txt-eng-ft.Plo uses 
> /usr/X11/include/fontconfig/fontconfig.h
> Please fix build process to get consistent use of fink's headers.
> ### execution of /tmp/fink.7On5c failed, exit code 255
>
> Any idea of how I might repair whatever is wrong with my fink 
> configuration on this computer?
>
> By the way, I was able to build octave-x11 on this machine, w/o 
> monkeying around with .info files.
>
> Thanks,
> Andy
I'm not sure yet.  The information that octave-x11 works but octave does 
not does at least help restrict the packages that _could_ be 
problematic.  Since the only real difference between octave and 
octave-x11 is which version of the FLTK library is used, I'm suspicious 
of fltk13-aqua.

-- 
Alexander Hansen, Ph.D.
Fink User Liaison
http://finkakh.wordpress.com/2012/02/21/got-job/


------------------------------------------------------------------------------
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here 
http://p.sf.net/sfu/sfd2d-msazure
_______________________________________________
Fink-users mailing list
[email protected]
List archive:
http://news.gmane.org/gmane.os.macosx.fink.user
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-users

Reply via email to