Re: [Axiom-developer] build-improvements and more-rules.mk

[C Y]

--- Waldek Hebisch <address@hidden> wrote:
> > > 
> > > I agree that algebra bootstrap should be revisited.  But IMHO the
> > > main work to do is outside Makefiles: I think that algebra should
> > > build just from spad files using Spad compiler and almost empty 
> > > databases.
> > 
> > Agree, but that's more than just the main work - it's a massive
> > undertaking.
> 
> I think it is easier then you think.  At the begining one has no
> idea what to do.  But once you start changing things the system
> break down and studing breakage teaches you a lot how it works.
> And regardles of how much work it takes it must be done: IMO system
> which has huge uneditable blobs as "sources" too hard to maintain.

Ah, I see what you're getting at.  (Sorry, the "just from spad" part
triggered the old boot language debate memory files.)  Yes, absolutely
- any maintainance nightmare we can remove from Axiom has to be
removed.

> AFAICS axiom_cmu build algebra the same way as GCL version. Main
> difference is that axiom_cmu makes some ANSI corrections (but not
> enough to compile it using SBCL), and skips boot directory.
> Motivated by axiom_cmu I made a patch to build-improvement branch
> which skips boot directory:
> 
>
http://lists.nongnu.org/archive/html/axiom-developer/2006-10/msg00039.html

Ah!  I remember seeing that but I confess I did not appreciate the full
significance of it.  Very nice indeed.

Gaby, is there any plan to incorporate this into the build-improvements
branch?  I guess it might be a bit out of scope for that branch...

Cheers, and thanks!

CY

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com