Re: [Axiom-developer] code release

[root]

>Great. I suspect this will make most of the build troubles go
>away. However, (just as a reminder) please call it BETA (or nothing at
>all) until the proposed fixes are integrated.

Umm, I don't propose to call it anything. I leave that up to the director
of marketing. I'm going to download, build, and test axiom--main--1 on
all the platforms I have available and, if it passes, promote it to 
savannah. Then I need to grab the savannah copy and redo the testing to
ensure I didn't make a mistake. Following that we have to chase down all
of the binaries and upgrade them. 

I don't yet have the ability to do this for the windows versions. We
need to get the scripts in place so I can automate the task but
probably not for this version as it is bad practice to change anything
at the last minute.

>Is it possible to "export" the list of bugs from savannah somehow?

By hand, I'd assume. Bug tracking has traditionally been done in the
input subdirectory. Several files there have bug fixes with listed
dates as well as some known bugs that are pre-NAG. Besides, these
bugs need to be better documented. I've been so busy making new bugs
and killing them before they get off my desk to concentrate on the
outstanding bugs.

>And since I still didn't get an answer: what about bug #5977 ? Do you have a
>patch lying around somewhere. If not, we really should "reopen" it!

It appears that the problem is a semantic one. The listed "bug" is
requesting that (x^2) be added to 1::DMP([x],INT) rather than
(x^2+1)::DMP([x],INT), which works.

==========================================================================
-> axiom
                        AXIOM Computer Algebra System 
              Version of Tuesday October 26, 2004 at 23:09:19 
-----------------------------------------------------------------------------
   Issue )copyright to view copyright notices.
   Issue )summary for a summary of useful system commands.
   Issue )quit to leave AXIOM and return to shell.
-----------------------------------------------------------------------------
 
(1) -> (x^2+1)::DMP([x],INT)

         2
   (1)  x  + 1
                         Type: DistributedMultivariatePolynomial([x],Integer)
==========================================================================


Now doing this:

==========================================================================
(2) -> 1::DMP([x],INT)
      

   (2)  1
                         Type: DistributedMultivariatePolynomial([x],Integer)
(3) -> %+x^2
 
   >> System error:
   Caught fatal error [memory may be damaged]

protected-symbol-warn called with (NIL)
==========================================================================

clearly fails but that's the color of a different duck and, clearly, is
still an outstanding bug.


>Another bug issue: The possibility of using axiom to demonstrate failures in
>IssueTracker is superb, I do agree. But caution: it is absolutely necessary to
>include a version number for each of those bugs, since, as soon as axiom is
>upgraded on MathAction, these bugs won't be reproducible there anymore!

IssueTracker? Did I miss a memo?

Tim