[Top][All Lists]

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

[Gcl-devel] Re: cvs commit ? PS

From: Camm Maguire
Subject: [Gcl-devel] Re: cvs commit ? PS
Date: 08 Jul 2004 17:09:08 -0400
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2


Michael Koehne <address@hidden> writes:

> Moin Camm,
> > Would you like to become a GCL developer?  If so, please register at
> > savannah and I can set you up.  This would be great!
>   just registered - think I'll have the web-robot-reply when finished
>   this eMail.

OK, you have been added at savannah.  Welcome aboard!

> > Your changes are among the most far reaching of any of the other
> > developers beside myself up to this point, which is a good thing!
>   I hope that there are some area's that I dont need to touch. Namely
>   garbage collection, loading of .o & .dll files and similar things
>   that are not necessary for an 'application level of ANSI conformance'.

OK, lets consider you an ansi compliance expert :-)!

> > Most other developers have either a special task (e.g. ansi tests), or
> > work primarily on a special port, so there is seldom any collision of
> > effort. 
>   lets say I'm trying to Port CLC with every Debian/CLC application
>   that does not require UFFI or Gray-Streams. Current status is still
>   18/124 applications, as pprint is not trivial to implement.

I'd like to write about FFi soon.

> > I'd like to start a discussion on gcl-devel on the goals and direction
> > of 2.7.0 in the near future, which should make all of this a lot
> > clearer.   For the time being, I'd suggest that we create a separate
> > branch for you to work on any of your changes which affect the current
> > build design (i.e. the cltl1/ansi layering scheme), and leave cvs
> > head/2.7.0, at least for the time being, with the cltl1 directories
> > 'bug fix only' to the extent possible and with new features placed in
> > separate directories added only in non-traditional builds.
>   the danger with two branches is, that the experimental branch could
>   miss bug fixes and improvements done in the main branch.

Would cvs merge suffice?  I often do something like cvs -z9 -q diff -u
-r Version_foo -D date1 -D date2 | patch -p0 -R [ --dry-run ]

> > So if this is acceptable, I'll take your patch, apply it as one whole
> > to CVS head as you request, test, move as much of the new stuff as I
> > can out of cltl1's way, retest, and then commit if all is well.  In a
> > separate branch, we can commit the patch as is 'unmoved', and you can
> > work here to integrate clcs into the core, rewrite pcl in C, and work
> > toward a small native ANSI build in current cltl1 size.  Needless to
> > say, if you could achieve this, it would be great!  Such a branch
> > would then be a candidate for a 2.8 release, which might even be
> > numbered 3.0.
>   I would prefer, if things like logical pathname translations could stay
>   in raw_pre_gcl, as they could simplify the build process, if I find
>   the time to rework it. There will be some ./mod/ansi_assert.lisp soon
>   in my patch, as ./lsp/gcl_assert.lsp does not play well with ./clcs/
>   This ./mod/ would be used INSTEAD gcl_assert during the raw_pre_gcl ->
>   saved_pre_gcl step, untill ./clcs/ provide their own macros for the
>   final saved_ansi_gcl.

OK, I'll try to look into it.  Alas, I've gotten a bit bogged down, so
Sunday looks the earliest now :-(.  As we discussed, I'll apply your
patch as a whole locally, move as necessary, and commit in one step.
We can discuss rearrangements in either head or a separate branch as
necessary.  Once this is done, perhaps we can informally designate
certain files on which you can serve as the local expert and hold the
main reigns on.  E.g. pathname.d comes to mind :-).

Take care,

> Bye Michael
> -- 
>   mailto:address@hidden             UNA:+.? 'CED+2+:::Linux:2.4.22'UNZ+1'
>   http://www.xml-edifact.org/           CETERUM CENSEO WINDOWS ESSE DELENDAM

Camm Maguire                                            address@hidden
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah

reply via email to

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