Re: EMACS Ada Mode Maintenance

[john@assen.demon.co.uk.nospam (John McCabe)]

dewar@gnat.com wrote:

>In article <79edut$cup$1@nnrp1.dejanews.com>,
>  briot@gnat.com wrote:
>
>  <<send your patches>>
>
>Just to clarify here: patches that are of general use for Ada, and
>relevant to its use on GNAT are definitely welcome. Patches that
>work ONLY for some proprietary compiler are contentious. I doubt
>Richard Stallman would be willing to approve such patches for
>inclusion in the official EMACS sources, so such patches are probably
>not appropriate.

I agree. Patches for other compilers should be forwarded for inclusion
into Elisp Archives, sent to gnu.emacs.sources, maintained by the
vendor and so on like is done with VC-Clearcase. (Unfortunately this
resulted in the version of VC-Clearcase that I got from Rod Whitby's
site being rather out of date with respect to the version of vc.el
that comes with the latest copy of Emacs, but I guess these things
happen!)

>So patches that are specific to some other compiler should probably be
>managed by the vendor of that compiler as an optional addition.

Or by individuals through Elisp archives etc...

>Note that with EMACS, such plug-in additions are of course easy to do,
>and do not typically require recompilation or reinstallation of EMACS.

Of course, which is the whole point of what the discussions I have
been having with yourself, Emmanuel, Rolf and Matt have been about
over the last few weeks i.e. that there was really no need for
ada-mode.el to be hacked around as much as it was to provide the
support you wanted for GNAT, it could have so easily been done by
extension.

>I am doing my best to interpret Richard Stallman's intent here (his
>actual language in private messages to me with respect to proprietary
>compilers was quite definite. I commented to him that there had been
>some discussions about support of other compilers, but they seem to
>be resolved, and he objected to any attempt to "placate" the users of
>other compilers, on the grounds that this had nothing to do with the
>GNU project).

I know I've mentioned users of other compilers in the past, but I
didn't intend this to suggest that they be supported, merely that the
functionality already provided by ada-mode.el that makes it easy to
use different compilers should not be removed from that file, only
extended.

>I think it is useful to keep this in perspective. Both EMACS and GNAT
>are part of the GNU project. It is thus quite natural that EMACS be
>specially aware of the requirements of GNAT.

Granted, but I consider ada-mode to be a facility to assist in
programming in the Ada language, not for programming with the GNAT
compiler. On this basis I would expect default setting should be for
GNAT commands, but I would hope that these could be provided through
configurable function calls.

>As I said in an earlier note, the GNU project is not going to go out
>of its way to make EMACS inconvenient to use for other compilers, but
>it certainly will not make special attempts in the other direction.

Again, all I would ask is that future development of ada-mode should
apply language specific features to ada-mode.el, with GNAT specific
features being provided by external packages, through mechanisms (a
kind of API so to speak) supplied in ada-mode.el where possible, 

>I hope this clarification is useful. There seemed to be some concerns
>in this thread earlier on that it was somehow inappropriate for GNAT
>to be creating a GNAT-specific version of Ada mode, and that this
>was not the official version and was therefore somehow unhelpful.
>Hopefully Richard Stallman's views on this matter are now clear, as
>well as ours.

The only thing I really felt was inappropriate was the use of the term
ada-mode to describe the package you created, rather than GNAT-mode.
Having looked at the code, I believe it would have been possible to
extend ada-mode through additional packages (gnat-mode??) to provide
the functionality you have done, *without* significantly changing the
structure of ada-mode.el.

>In any case, I think in practice the outcome of this development will
>be continued improvement of Ada mode for EMACS that will benefit all
>users of this very useful technology, and we welcome contributions.

I think it has certainly been worth my while 'sticking my oar in' so
to speak as it looks like the future of ada-mode is pretty secure.

>Emmanuel will shortly be posting an announcement here reminding people
>about the EMACS Ada mode mailing list, and will also be releasing an
>updated version of the EMACS Ada mode sources in the very near future
>(his plan is to have frequent updates and releases of this technology).

Sounds good.

I'd personally like to thank ACT again for providing the resources
necessary for Emmanual to do the job.


Best Regards
John McCabe <john@assen.demon.co.uk>