axiom-developer
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Axiom-developer] Re: root chunks


From: C Y
Subject: Re: [Axiom-developer] Re: root chunks
Date: Fri, 22 Jun 2007 11:28:04 -0700 (PDT)

--- Stephen Wilson <address@hidden> wrote:

> Hi Cliff,
> 
> C Y <address@hidden> writes:
> > Well, this being Lisp, I could just program in more power, set
> > working defaults, and let the user deside.  Default Lisp and boot
> > files to one behavior, spad to whatever the algebra needs, and
> > have some variables to switch behavior around if anyone needs 
> > it :-).
> 
> This is basicly the approach I would take.  The system should have a
> surface simplicity for the user which needs no more,  but facilities
> available for the developer who needs fine grained control.

Can do.

> Apologies that this comment does not help much in achieving that
> goal.

No problem.  It's not in a working state yet - I think my position at
this stage is basically that I know how to achieve pretty much what I
want, but I need to wade through the nitty gritty of doing it.  The
most vexing annoyance is probably the bootstrapping of the boot
language - that will probably take some special operations and tricks
to do the way I want to.  I'm doing something a bit different from
"normal" asdf asd files - there are a couple commands other than system
definitions (I can't do literate defsystem definitions without
asdf-literate loaded) and I'm defining all the "systems" in one
pamphlet file - that's because I think it is logical that
axiom-system.asd.pamphlet discuss the overall system design and
component interactions for the human as well as asdf.

> However, I do have plans to look at your work in detail once a
> few more items get taken care of.

I plan to show it to the world once I reach the point where I can
successfully handle lisp, boot and spad pamphlets.  Other goodies like
the hyperdoc system and c files for sman and friends will have to come
later.  I'd like to look at alternate solutions to some of that anyway
but that's down the road (might as well not annoy folks with more Lisp
talk.)

> BTW, have you tried ASDF with GCL lately?  I have a slightly hacked
> version which basicly works with 2.6.8pre in ANSI mode.  In the sort
> term I hope to be switching my work to be based on 2.7.0 cvs, and I
> expect ASDF to be a fair bit more comfortable there.  The notion of
> integrating cl-web with asdf is fantastic.

:-).  I've got just enough working that it's getting fun.  I'm working
in SBCL at the moment - once I have something working we can look at
porting it to GCL et. al.  I'm not too interested in supporting the
legacy GCL - 2.7.0 makes more sense to me.  It's enough effort to do
all this without having to worry about backwards compatibility - I
would suggest we release a gold version soon and make that the last
gold version based on a non ANSI lisp.  Based on attempts so far I'm
afraid merging the ANSI changes into Silver may be an ordeal, but it is
a necessary step.

I'm using files from wh-sandbox right now since a) its got the most
stuff working and b) it works with sbcl.  Probably a necessary step
will be to find the changes relative to the Silver branch that make
ANSI work and submit them as an offering to Silver - without that
folding this back into Silver probably won't be possible (or at least,
a big pain).

Cheers,
CY


      
____________________________________________________________________________________
Fussy? Opinionated? Impossible to please? Perfect.  Join Yahoo!'s user panel 
and lay it on us. http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7 





reply via email to

[Prev in Thread] Current Thread [Next in Thread]