Re: [Axiom-developer] NAG libraries

[Stephen Wilson]

Waldek Hebisch <address@hidden> writes:

> No, I erased the corresponding tree. 

Oh well, the process appears fairly straight forward.  

> ATM I am not sure what to do with NAG library support.  Large part
> of it is very specific to NAG library but other part have some
> potential for reuse.  Also, even if we do not want to use some parts
> it may still serve as a template showing how to solve some problems.

My perspective on the issue is that as the code is basicly unused at
the moment, it can go.  The code lives in the SCM history (among other
places) and can be retrieved at any time.  Any role it might serve as
template for future work will not disappear.

> I belive that you can remove the following algebra files:
> 
> annacat.spad     d01routine.spad    d02routine.spad  e04agents.spad
> compile.input    d01transform.spad  d03Package.spad  e04routine.spad
> cont.spad        d01weights.spad    d03agents.spad   functions.spad
> d01Package.spad  d02Package.spad    d03routine.spad  routines.spad
> d01agents.spad   d02agents.spad     e04Package.spad  tools.spad
> 
> asp.spad  d01.spad  e02.spad  f04.spad      fortpak.spad
> c02.spad  d02.spad  e04.spad  f07.spad      fortran.spad
> c05.spad  d03.spad  f01.spad  fortcat.spad  forttyp.spad
> c06.spad  e01.spad  f02.spad  fortmac.spad  s.spad
> 
> On Nag cdrom the first group is in 'anna' subdirectory the second
> is in 'naglink2'.  IIRC the rest of algebra does not depend on
> them.
> 
> We also have a lot of Nag support in interp subdirectory --  at least
> nag* and anna.boot.pamphlet

Thanks for the list! It appears to be a slight super set of the list I
had written down.  I will investigate and likely remove all mentioned
files.

Ill post the result and at least make this a feature of the axisp
branch.  It would seem better to wait until a more general mechanism
is hammered out before proposing the same for silver.


Thanks again,
Steve