Re: [lawrence mitchell] Re: [PATCH] erc: make erc-generate-log-file-name-long and erc-log-file-coding-system xemacs-compatible

>>>>> "Stephen" == Stephen J Turnbull <[email protected]> writes:

    Stephen> Sorry about the repost ("fat fingers on C-c").
    APA> what is the right way to specify a catch-all coding system in
    APA> Mule and non-Mule XEmacs, supporting XEmacs 21.1 and later?

    Stephen> What do you mean by a "catch-all coding system"?

A quote from
http://cvs.sourceforge.net/viewcvs.py/erc/erc/erc-log.el?r1=1.7&r2=1.8

Look for Line 153 in the new revision.

    >>> How about something like:

    >>> (if (featurep 'mule)
    >>> 'emacs-mule ; or some other catchall
    >>> 'buffer-file-coding-system)

    Stephen> This is definitely wrong.  'emacs-mule doesn't exist in
    Stephen> XEmacs (this is the internal coding used by Mule, and
    Stephen> only a very few other applications support it), and
    Stephen> coding systems don't exist at all in non-file-coding
    Stephen> XEmacs.  That means almost all non-mule installations on
    Stephen> Unix.  I would guess that any place where coding systems
    Stephen> are used will have to be rewritten to test for (featurep
    Stephen> 'mule) instead of trying to use a generic coding system.

    Stephen> If you don't like that, I'd say just require file-coding.
    Stephen> Let's ask Vin how he feels about it, maybe we should
    Stephen> think about moving to defaulting to file-coding in 21.4.

Vin?

-- 
Adrian Aichner
 mailto:[email protected]
 http://www.xemacs.org/