Hello Dave !!!

I feel I can give bad answers to each of your question, as I may be at
fault for each of them :-)

1) 2) If you think these options may be of help, there is no reason
why they shouldn't be added. I don't actually remember to have looked
at the list of them when I struggled with this spkg.

3) I definitely remember having had problems to get GLPK to compile as
a SPKG, and I wouldn't be surprised if it did not without this
library.

4) Indeed, this is totally useless if GLPK is made standard. I could
also remove some code from mip.pyx which ensures that "at least one
solver is installed"

5) Do you mean that it works when you run "./configure && make"
manually, or as a SPKG ? It would mean I was wrong in my answer to 3.
I don't remember the bug to be platform-dependent.

6) The answer to the "c++" may be that I was working in parallel with
the CBC package, which requires C++, and I tried a lot of stupid
things to debug it. I think it can be safely removed, as I don't
recall seeing any C++ code in GLPK ! About the path : $SAGE_LOCAL is
for me a bash variable which can not be used in this file, and I
wonder if the Sage variables defining these paths are defined at this
point. If anyone reading this knows how to access them... :-)

7) You are right about the patch/ directory. The point is that early
version of the SPKG required Sage to use this external code, and
someone who did download Sage during this period still requires the
GLPK spkg to contain these files. WHen I moved this code to the
standard Sage library, I thought it better to let it sty in the SPKG
for a while, just for those people, and this is the only reason. It
has been som time already, by the way, so this may indeed be a good
time to get rid of them.

8) YEEEEEEEEEEES ! Since this spkg is available in Sage, quite some
patches have been based on it. You will find any occurences of
"#optional - GLPK" in graph.py, generic_graph.py digraph.py (graph/
foler) or in almost all the files of the numerical/ folder.

Thank youuuuuuuuuuuuuuu :-)

Nathann

-- 
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to 
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org

Reply via email to