On Feb 12, 8:01 am, mabshoff <mabsh...@googlemail.com> wrote:
> On Feb 12, 7:50 am, kcrisman <kcris...@gmail.com> wrote:
>
> <SNIP>
> .
>
> > At the very least we know Sage has its work cut out for it if it ever
> > wants to remove dependence on the slow-slow interface to Maxima and
> > Lisp issues, because these are (in general) very thorny questions.
> > Even if they're amusing on occasion!
>
> Yes, getting to the point where Maxima is today will be a formidable
> amount of work.
>
> The issue I have with Maxima is that back in the day Macsyma was
> designed under the assumption that someone would always be sitting in
> front of the terminal and give feedback, i.e. is answer questions
> whether something is "positive, zero or negative".
Exploring alternatives to this is an active area of research, and has
been for at least 20 years.
There are several unsatisfactory alternatives.
> This is highly
> annoying and a major weakness IMHO, i.e. the latest maxima will ask
> you about a when you run
>
> limit(a*x,x,0);
>
> Why? a's sign has nothing to do with the limit here. Maxima 5.16.3
> will not ask you about the sign of a in the above limit.
Presumably this is a bug, not a design issue.
> That is the
> reason why we did not upgrade to Maxima 5.17.1 because it introduced
> more cases where user interaction was required.
This behavior is presumably a bug. Rejecting all bug fixes in the
latest version of
Maxima is, perhaps, an error.
> If one wants to run
> Sage computations involving Maxima one does not want to answer
> questions.
I think this is wrong. I think you do not want to answer questions
that you
perceive to be unnecessary.
And no one argues with that.
There are questions that you probably DO
want to see. Like "I've run out of memory page space. I can continue
but
will run at disk speed, 10,000 times slower. Do you want to restart
with a
larger memory allocation?"
"If there is a possibility that <some expression> is zero, then some
<expensive extra computation>
should be done. Should I do so?"
>On top of that the assumption system in Maxima is woefully
> under documented, i.e. stripped of comments,
There is only one case in which an author removed comments (circa
1971)
from his code on the grounds that anyone looking at the code with
comments
might be tempted to change the code. So he removed his comments.
That code had to do with the assume database, which is now considered
to be a source of problems.
The lack of comments in other programs is just the way they were
written.
> and only in the last
> months or so Dieter Kaiser has started to fix long standing bugs in
> it. I am always surprised given the age of the code base how many
> issues are left in Maxima and how often code is broken. I.e. in 5.13
> or 14 some limit involving exp(-x) was broken.
It is unfortunate that new bugs are introduced. Open source is tough
that way.
As for how many issues are left in Maxima --- well, there are design
issues
that persist in every computer algebra system. Mathematics is like
that
sometimes.
>
> > - kcrisman
>
> Cheers,
>
> Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
