Re: Init error message

[David Kastrup]

Let me add as a prerequisite that I have no talents to be polite or
diplomatic.  So don't take posts like this personal.  I have
Xpost/Fuped this to comp.emacs and comp.emacs.xemacs which I consider
more fitting to this discussion (even though it might be considered
sufficiently rehashed to death) than gnu.emacs.help.

John Paul Wallington <jpw@gnu.org> writes:

> Le Wang <lewang(at@)yahoo.com> wrote:
> 
> > It's in my (and our) interest to see both efforts continue and
> > thrive.
> 
> This is a matter of opinion.  See for example:
> http://groups.google.com/groups?q=&selm=rjd6z1x1qt.fsf%40ssv2.dina.kvl.dk
> 
> Do you think a merge is desirable?

Yes.

> Do you agree that in order to merge the projects, one of the
> projects has to be shutdown?

Yes.  There _is_ already merging going on all the time in the
direction Emacs->XEmacs.  This does not cut it for application
developers.

It is clear that the survivor must be GNU Emacs copyrighted by the
FSF.  Emacs is a core piece of GNU software, the FSF will not
relinquish control, Stallman himself is actively working on it.  This
does not mean that core pieces of XEmacs could or should not be
integrated into GNU Emacs and replace stuff there that is suboptimal,
but it means that integrators would need to be willing and legally
capable of signing over their work to the FSF, _and_ would need to get
along well enough with Stallman in order to progress with substantial
changes to a core project of his, including extensive changes and
documentation of internals.

Stallman is a very headstrong person.  While he is not malignant, it
can be a chore working with him.  Free Software needs a man of his
qualities, so I tend to arrange what it is necessary in order to
cooperate with him, even though there are times when he drives me up
the wall.

I don't see too much willingness of the _core_ developers of XEmacs to
cope with what would be necessary for a merge.  And frankly,
cooperating with Stallman on a comtinued basis is not something that
one could reasonably _demand_ from anybody to cope with.  If at all,
such participation would have to be voluntary, and on a project as
large as a merge of XEmacs and Emacs it would be very taxing to both
parties.

Personally, I think that one way to improve matters would be that if
XEmacs developers came up with some new functionality and/or API, they
would discuss it also on the Emacs developer lists and would offer to
sign the respective papers and stuff in order to help implementing it
also in Emacs.  And if the original plans get thwarted for some
different interface of comparable usefulness, rather implement that in
XEmacs.

Yes, it will expose them to GNU Emacs and Stallman more than they
like, but it will help lessen the amount in which application
programmers get screwed that have to cater for both Emacsen.

This will not buy us a unified Emacs, but it would lessen the bad
impact of the split.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum