On Jun 29, 2008, at 6:54 PM, David Joyner wrote:

> Thanks but although that eliminated one traceback error, it created  
> another.
> Also, I'm worried that hacking Mike Hansen's combinatorial_algebra  
> module
> will create much more serious problems in other parts of SAGE.

The apparent inability of CombinatorialAlgebra to mesh well with the  
rest of Sage seems a defect in CombinatorialAlgebra that will  
hopefully be fixed (looks like Mike Hansen is on it, perhaps it's  
already done in the sage-combinat branch). From what I've read of it,  
this module seems well written, so the changes shouldn't be too  
invasive.

> I'm starting to think that either GroupRing should be written from  
> scratch
> (based closely on CombinatorialAlgebra)

This will create a huge amount of redundant code.

> or else a GAP wrapper should be written.

Wrapping something so basic in GAP has several disadvantages--then  
group rings would be limited to groups and rings that GAP understands  
(which, especially on the rings side of things, is much more  
limited). There's also the issue of speed--when permutation group  
elements were converted from a GAP wrapper to a native implementation  
there was a 4400x speedup (mostly due to overhead).

> I'm mostly interested in order to implement David Kohel's split  
> group codes,
> which generalize the duadic codes which have recently been  
> implemented in
> SAGE. Unfortunately, they rely heavily on abelian groups and their  
> dual,
> which is currently in a state of flux (I think David Roe is rewriting
> it). So maybe this should wait?

I am generally against waiting for things... hopefully the interface  
will still be (mostly) backwards compatible through the rewrite.

- Robert


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/sage-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Reply via email to