On Sun, 13 Jan 2013 20:28:59 -0500, Daniel Macks <[email protected]> wrote:
On Sun, 13 Jan 2013 16:24:43 -0700, Alexander Hansen
> <[email protected]> wrote:
> Putting this on the list, because this is not MY package. >
> > On 1/13/13 11:31 AM, Shih-Kai Chou wrote:
> > > > So far, I think my problem is not fixed. (no one mail to me)
> >
> > I didn't reproduce your problem, no. >
> > > I tried to realized it on my own way although I am not good at Linux and
> > > programming language. > > > Because It says : > > cd data/ ( the
> path is > /sw/lib/ccp4-6.2.0//data ) (In your case, I think
> > > it is ccp4-6.3.0 ?)
> > > ./fontpack
> > > At line 12 ...... > > > I just cd there, and execute it
> (./fontpack) to see what is information
> > > it will give me!
> > > > >
> ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
> > > At line 14 of file
> > > /sw/src/fink.build/ccp4-6.2.0-102/ccp4-6.2.0/lib/data/fontpack.f (unit =
> > > 11, file = 'font84.dat')
> > > Fortran runtime error: Cannot write to file opened for READ
> >
> >
> > It _shouldn't_ say
> >
> "/sw/src/fink.build/ccp4-6.2.0-102/ccp4-6.2.0/lib/data/fontpack.f"--apparently
> > this got set during the build and probably needs to be fixed. >
> > However, this probably isn't relevant, since you're comparing files from
> > two different versions of the package. I don't know anything about
> this package or its confusing build system, but I do see two places
> in ccp4.info that do hard-code non-constant paths. line 298: perl
> -pi.bak -e 's|/src/fink.build/ccp4-%v-%r|/opt|g' bin/clipper-config
>
> line 307: perl -pi.bak -e
> 's|/src/fink.build/ccp4-%v-%r/ccp4-%v/lib|/lib/ccp4-%v|g' lib/**/*.la
>
> It is not valid to assume the path to the build-dir (the
> "src/fink.build" part there). I assume those lines have the fink
> prefix first ("/sw/src/fink.build" for example), so one should be
> using something related to the %b token. If one is hacking .la files
> that are actually generated by libtool (rather than a custom-written
> script that simulates some features of it) and you're getting
> build-dir paths in it, the *actual* problem is in the makefiles or
> use of libtool. Modern libtool knows the difference between "build
> dir" and "eventual runtime dir" and especially knows to use the
> latter for .la.
Couldn't even get that far. First, missing BuildDepends:libiconv-dev
(no big deal). Then 6K lines of linker errors (no, seriously!) because
I'm on 10.6/i386 and the .info hardcodes /usr/bin/{gcc,g++} which use
the x86_64 arch instead. Need to use just "gcc" and "g++" because fink
does some behind-the-scenes magic to make the compiler use the "right"
arch for how the user has fink configured, whereas hardcoding uses
whatever apple decides is best this week.
dan
--
Daniel Macks
[email protected]
------------------------------------------------------------------------------
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
MVPs and experts. SALE $99.99 this month only -- learn more at:
http://p.sf.net/sfu/learnmore_122412
_______________________________________________
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
