Re: CVS commit: pkgsrc/graphics/fotoxx

[Ryo ONODERA <ryo_on%yk.rim.or.jp@localhost>]

Hi,

From: Bernd Ernesti <netbsd%lists.veego.de@localhost>, Date: Fri, 24 Aug 2012 
20:47:10 +0200

> On Fri, Aug 24, 2012 at 07:06:48AM +0000, David Holland wrote:
>> On Fri, Aug 24, 2012 at 07:19:03AM +0200, Bernd Ernesti wrote:
>>  > > zfuncs.h:42:22: fatal error: execinfo.h: No such file or directory
>>  > > 
>>  > > I guess this is because the -current on my test-build machine isn't
>>  > > new enough. Is it worth improving the conditional with something like
>>  > > ".if exists(/usr/include/execinfo.h)"?
>>  > 
>>  > No, the version check has to be fixed. Right now it is 699000000.
>>  > 
>>  > 699000800 would be better because that is after it was added in May.
>> 
>> But that kind of check is a major nuisance in make. What's wrong with
>> testing for the header inside a check for NetBSD?
> 
> Because I only checked the patch file and didn't notice that the Makefile
> has besides a pkgbump also a check.
> 
> Btw, the new Makefile now asumes that every system who has a
> /usr/include/execinfo.h has a libexecinfo too.
> Is that true for Linux which was the only supported system before?
> And is that check ok with cross builds?

I have checked only on NetBSD 5.1 and current.
As far as I can recall collectly, on Linux, backtrace(3) is part of
glibc.
I will fix Makefile of fotoxx.

Thank you.

>> ...maybe it's time we taught make how to do pkgsrc-Dewey.
> 
> ?

Me too.

--
Ryo ONODERA // ryo_on%yk.rim.or.jp@localhost
PGP fingerprint = 82A2 DC91 76E0 A10A 8ABB  FD1B F404 27FA C7D1 15F3