Missing CHOLMOD in octave 2.9.9

David MacMahon davidm at astro.berkeley.edu
Tue Feb 27 11:12:47 PST 2007


On Feb 27, 2007, at 10:48 , Andre Stechert wrote:
> Sorry, didn't realize you had work in progress :-(.

Sorry, I didn't tell anyone! :-( :-)

> It's easy enough to remove the curl reference.

Will do,

> I also made the patches necessary to fix glpk, as per the  
> discussion with
> Andrew Makhorin here:
>
> 	http://www.nabble.com/Updating-glpk-binding-to-GLPK-4.15- 
> t3250570.html
>

I've removed my parallel work from the upcoming patch.

> Which leaves the following reports:
>
> 1) some folks are running into a problem with one of the libs not  
> being ranlib'd.
>    the folks that i talked to were running on panther.  no fix for  
> this yet, plus
>    haven't reproduced it.

Didn't know about that one.

> 2) some folks seem to have a version of the fnmatch header around  
> that's not smart
>    enough to put extern "C" around the fnmatch declaration (or the  
> dependent sw
>    is declaring it itself).  also haven't reproduced this.

If they are somehow getting /usr/include/fnmatch.h to compile with a  
compiler that does not define __cplusplus, then they will not get the  
extern "C".  Perhaps octave's config.log from someone who encounters  
this problem would be informative.


> 3) octave/defaults.h contains wrong paths after installation  
> (blocking octave-forge
>    port) http://trac.macosforge.org/projects/macports/ticket/ 
> 11181.  Do you have a
>    fix for this, given your work on octave-forge?
>
> 4) destroot.destdir keyword unnecessary
>    http://trac.macosforge.org/projects/macports/ticket/11167

These two are fixed in my upcoming patch.  They are really the same  
thing.  #11167 nailed it on the head,

Dave






More information about the macports-users mailing list