[Maxima-commits] [git] Maxima CAS branch, master, updated. branch-5_30-base-152-g7360758

Am 30.07.2013 17:53, schrieb Raymond Toy:
> On Tue, Jul 30, 2013 at 7:18 AM, Volker van Nek <
> van_nek at users.sourceforge.net> wrote:
> 
>>
>>
>> - Log -----------------------------------------------------------------
>> commit 7360758b26cd7f441ce8f6100a1474dd7e31542c
>> Author: Volker van Nek <volkervannek at gmail.com>
>> Date:   Tue Jul 30 16:18:05 2013 +0200
>>
>>     update files in contrib gf
>>
>> diff --git a/share/contrib/gf/aes.mac b/share/contrib/gf/aes.mac
>> index 50c0f35..9536c52 100644
>> --- a/share/contrib/gf/aes.mac
>> +++ b/share/contrib/gf/aes.mac
>> @@ -34,7 +34,7 @@
>>
>>     All functions are compiled when loading aes.mac. Takes a few seconds.
>>
>> -   Functions in this file work in Maxima version 5.29 or higher.
>> +   Functions in this file work in Maxima version 5.31 or higher.
>>
> 
> Does this mean this code won't work in any version of maxima earlier than
> 5.31?  Seems unfortunate, but I guess users can use an older version for
> maxima 5.30 and earlier.
> 
> Ray
> 

Ray, users of Maxima 5.30 have their own version of aes.mac in
share/contrib/gf.

The difference between the 5.30 and 5.31 version of aes.mac is that I
changed the way of computing a matrix inverse in gf from the adjoint
method to LU decomposition and as a consequence I changed the function
name from gf_matinv to gf_invert_by_lu. Calling gf_matinv still works in
5.31 but prints the information that it is deprecated.

I revised the gf package in large parts in the last months and added a
lot of new stuff. I believe that it is now in some kind of final stadium
and I am going to write the texi documentation in autumn. At this point
I think all function names will be fixed.

Volker