[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Patch against latest G-Wrap
From: |
Andreas Rottmann |
Subject: |
Re: Patch against latest G-Wrap |
Date: |
Wed, 14 Sep 2005 21:36:24 +0200 |
User-agent: |
Gnus/5.1002 (Gnus v5.10.2) Emacs/21.4 (gnu/linux) |
address@hidden (Ludovic Courtès) writes:
> Hello,
>
> address@hidden (Ludovic Courtès) writes:
>
>> G-Wrap in address@hidden/g-wrap--dev--0--patch-18' is deeply
>> broken (doesn't compile). Below is a patch that fixes some of the
>> problems I encountered.
>
> I finally got to dig this further and a complete fix is available in my
> branch[*]:
>
> address@hidden/g-wrap--devo--1.9.6 (`patch-1' and `patch-2')
>
> The problems I had almost all related to the introduction of the new
> `(g-wrap c-codegen)' module. Previously, all the generation methods
> (`global-declarations-cg', etc.) where defined and exported by
> `(g-wrap)' but are now defined and exported by this new module.
> Consequently, modules that overload these methods must make sure that
> they use `c-codegen'. Failing to do so, these modules end up creating
> *new*, local, generic functions that are different from the ones
> exported by `c-codegen'.
>
Yeah, that code was utterly unfinished.
> Therefore, the next release will _have to_ mention this very clearly
> since it affects user code directly in an *incompatible* way (except for
> those who use the 1.3 compatibility API).
>
> Additionally, I removed `class-name' from the list of exports of
> `(g-wrap)', since this is unneeded and it triggers annoying warnings
> with Guile 1.7. I also changed the generation of the modules themselves
> so that `#:export' is used instead of the `module-use!' trick (see
> http://lists.nongnu.org/archive/html/g-wrap-dev/2005-08/msg00001.html).
>
> And the cool thing is that everything works fine, including the test
> suite! :-)
>
> BTW, I'm glad to see Andreas is working on a Scheme48 back-end.
>
Thanks for participating in development of G-Wrap! Note however, that
I can not promise that the changes done in the current --dev--0 will
lead to a release soon; all bugfixing will be done on the --dev--1.9
branch which will lead to a 1.10 stable series (maybe with the next
release). This will also mean that the API break you mentioned above
will happen at the start of the 1.11 unstable series. Considering I
seem to have co-developers now, I should probably announce such things
properly ;).
Cheers, Rotty
--
Andreas Rottmann | address@hidden | address@hidden | address@hidden
http://yi.org/rotty | GnuPG Key: http://yi.org/rotty/gpg.asc
Fingerprint | DFB4 4EB4 78A4 5EEE 6219 F228 F92F CFC5 01FD 5B62
v2sw7MYChw5pr5OFma7u7Lw2m5g/l7Di6e6t5BSb7en6g3/5HZa2Xs6MSr1/2p7 hackerkey.com
I have a truly elegant proof of the above, but it is too long to
fit into this .signature file.