Re: adding new algebra, was: Re: [Axiom-developer] typo in the guessing package pamphlet, was: Re: Algebra and literate documentation

[Bill Page]

On 08 Aug 2007 01:50:45 +0200, Martin Rubey wrote:
> Bill Page writes:
> ...
> > This is the same problem that was originally reported by Alfredo
> > Portes about this time last year.
>
> Yes.  I recall describing on the list how to work around it by making the 
> build
> process semi-automatic, i.e., adding only the files in the first layer (i.e.,
> RECOP, FAMR2, GOPT, etc.) to src/algebra/Makefile.pamphlet, finishing the
> build, then adding the files in the second layer to Makefile.pamphlet, and
> starting a new build, but specifying DAASE with
>
> make DAASE=forgot/the/path/sorry
>
> Too tired to look for the appropriate messages.
>
> In any case, this semi-automatic build is stupid and can be avoided by using 
> an
> appropriate build scheme.  I guess, Tim is working on that.  Really, the
> problem is the same as Franz Lehner's, I guess.
>

It is a similar problem but Franz' problem is even more difficult
because it attempts to make a basic change to some existing algebra
code deep down in the algebra hierachy. In this case the initial
databases that are in the share directory of the build can not be used
to compile this code. Thus there is no way to get the required
bootstrap started. The only version of Axiom that can currently
compile it is FriCAS since it has a more elaborate bootstrap system
that starts with empy databases.

Regards,
Bill Page.