Re: [Axiom-developer] stable release schedule

[C Y]

--- Martin Rubey <address@hidden> wrote:

> C Y writes:
>  >
>  > --- root <address@hidden> wrote:
>  
> I think that this is a good idea. There is one thing which should go
> along with this schedule however: *what* we want to target at. 

Makes sense.

> Whether the period will be one or three months does not seem so 
> important to me. We should, however, take into account major 
> holidays: Christmas, end of term in January (or is that an
> Austriacism?), Easter, end of term in June...

In the US end of term is usually a week before Christmas.

> So, to fill this with a little life, I'd like to have the first
> "real" release just before February with
> 
> * all the known (build and algebra) patches integrated
>   By the way: whats the status of bug #5977 x^2+1::DMP([x],INT) 
> 
> * my guessing package ready and documented. (that's my part, however,
> this shouldn't really be part of the distribution, should it?

Unless there are legal reasons to to include it, or it takes massive
amounts of space, I think it's easier for the end user to have
everything already there.  Pretty much all available Maxima code (with
one exception I'm aware of) is now included in cvs.

(Caveot - having high speed internet and a (relatively) good sized hard
disk, making a file a meg or two larger doesn't tend to faze me at all.
 This is not universally true, so I guess the question is who our
audience is and what would best serve their needs.  I would be very
interested if, for this first release, both Axiom itself and
Axiom+TeXmacs were packaged as stand-alone exe installs and we could
see which would be more popular.)

> While we are at it, here's another idea: I think it would be sensible
> to divide Axiom into two pieces: the Algebra and the rest. If it
> would be that way, we could have several tla/arch branches like 
> operating systems, book, graphics, hyperdoc, various lisp 
> implementations and one common branch algebra.

Um.  Shouldn't the code be modular enough to make this unnecessary?

> This way, the mathematics would be roughly the same for all branches
> and we wouldn't be in the awkward situation that Bill says "well, 
> I'll try to integrate some of the patches into the windows version 
> right now, but the linux versions have to wait until the weekend". 
> They would be the same automatically!

Oh, I get it.  Tim, you're probably the one to best answer this - isn't
the windows branch temporary until the main codebase can build on all
branches without trouble?  Or are there Windows changes that cannot be
folded back into main?  In that case, I agree with Martin.

>  > I like the idea (not that I'm qualified to have an opinion)
> 
> Why wouldn't you be qualified?

I've never had to coordinate something like that.  Maxima's release
schedule is basically "when we're ready" and anybody who wants to look
at the current state of things can check out cvs.  Also, history
teaches that whenever we put together a release there tend to be
non-math issues that pop up.  But if Axiom is checked regularly (i.e.
once every few weeks) on all platforms, perhaps there won't be too many
issues with making periodic releases.  That's what I'm not sure about.

CY


                
__________________________________ 
Do you Yahoo!? 
Send holiday email and support a worthy cause. Do good. 
http://celebrity.mail.yahoo.com