Le 05/07/2013 20:09, Volker Braun a écrit :
On Thursday, July 4, 2013 2:26:02 PM UTC-4, Snark wrote:
(1) My debian package only ships the dimension 11-enabled executables ;
Just to confirm, that is useless. Palp needs to be complied for
different dimensions.
Ok.
And you need to be careful with filesystem
timestamp granularity if you want to do that programmatically, see the
palp spkg.
I don't understand what you mean at all. I modified my debian package
(it's not uploaded yet) to do like the sage spkg (ie: dimension 4, 5, 6
and 11 with a default of 6, with the same executable names). That makes
all PALP doctests in sage pass.
(2) The current doctests detect that the default executable is compiled
with a higher POLY_Dmax, but do not detect that the executables return
wrong results below, so what you say means sage's doctesting should
definitely be improved too!
The real test is of course that one can access the 4-d reflexive
polytopes database, though thats a pretty big spkg to download for the
test ;-)
That raises two questions:
(1) Wouldn't creating just one graph by hand, suitably chosen so a
computation fails if the wrong POLY_Dmax-enabled palp is used, be possible?
(2) Shouldn't palp upstream detect that it is doing something wrong!?
In any case, upstream does not install multiple compiled versions (and
autotools doesn't really support that either). For better or worse, Sage
will have to build its own palp for the forseeable future.
I beg to disagree: the next upload of a palp package in debian will be
sage-compatible.
Snark on #sagemath
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/groups/opt_out.
