--- James Amundson wrote:
> CVS head will no longer build from scratch. "set" is still included
> in maxima.system, so build will only work if you have a compiled
> version of set left over.
Sorry for the bother, that was an oversight on my part.
I did execute make clean; make but clisp is more forgiving --
there was a warning message about the missing file but I didn't
see it, and the build did succeed. I have removed "set" from
maxima.system and committed it.
> I am perfectly happy to replace set with nset. However, first we have
> to
> 1) Cleanly remove set.
> 2) Take nset out of contrib and put it in the main distribution.
> 3) Fix any conflicts nset has with the rest of maxima.
> 4) Add tests for nset.
>
> I see that you have almost accomplished (1) and (4).
I hope (1) is accomplished now. I guess (4) depends on
resolution of conflicts with rtest15.
> (2) is important -- the code in contrib is unsupported by definition.
> Should we make nset part of the compiled maxima image?
I'm in favor of this. Maybe Barton and Stavros, the authors
of nset, want to comment on this.
> As far as (3) goes, I see at least one conflict: nset's definition of
> "every" is clashes with the "every" in mutils.lisp and the some of
> the tests in rtest15.
Hmm, when rtestnset runs, I see a warning about redefining
$FLATTEN (src/flatten.lisp) and $EVERY (src/mutils.lisp).
What is the conflict with rtest15 ?
Sorry again for causing a problem, I hope it's straightened
out now. Best,
Robert Dodier
__________________________________
Do you Yahoo!?
Yahoo! Mail - 250MB free storage. Do more. Manage less.
http://info.mail.yahoo.com/mail_250