[Axiom-developer] Re: freeing Aldor (was: closing bug reports)

[Martin Rubey]

Dear Bill,

"Page, Bill" <address@hidden> writes:

> Martin, 
> 
> On Friday, October 27, 2006 4:42 PM you wrote:
> > ... 
> > BTW, I think that issuetracker really needs one improvement: 
> > we need to be able to run old versions of axiom there. Many
> > many of the bug reports are strange to read, because the
> > examples provided don't fail anymore.
> >
> 
> Agreed. I have not found any easy way to do this nor enough
> time to do the necessary programming.

I think that this should not be so difficult. Is it possible for you to define
an environment like \begin{axiom}[optarg] ... \end{axiom} that takes an
optional argument which is passed to the script which calls axiom at the end?

You could simply install all patches of axiom we have in directories

axiom1, axiom2, ..., axiom50, ...

plus a symbolic link axiom which points to the last axiom which currently
exists - i.e., currently axiom50.

(Do we have enough space for so many axioms? Note that you can delete quite a
bit from these directories: at least all of doc and src)

Currently, your phyton stuff simply calls axiom and reads the collected stuff
within the environments of one page as an input file, doesn't it?

Now there are two possibilities: it could just call axiom in the right axiom
directory, depending on the argument of the first axiom environment of the page
- and simply ignore arguments of later axiom environments.

Or, it could collect inputs from environments with equal arguments, then
calling the corresponding axiom.

Maybe you could explain what the difficult part is. Maybe I get something
working then. Mind, though, that I don't know phython at all.

> Anyway, if someone would like to volunteer to lead an effort to contact these
> people formerly (and some currently) associated with Aldor, I think that
> would be much better than just remaining depressed. :-)

OK, I'll write something on Monday and send it to the list as proposal.

Martin