Re: Coding system of compressed PO files is not recognized

[Eli Zaretskii]

> From: Kenichi Handa <address@hidden>
> Date: Wed, 03 May 2006 11:46:55 +0900
> Cc: address@hidden, address@hidden
> 
> In article <address@hidden>, Richard Stallman <address@hidden> writes:
> 
> > I have a new idea.  We could establish the convention that you can
> > pass a buffer to find-operation-coding-system instead of a file name,
> > for the `insert-file-contents' operation.  And elements of
> > Vfile_coding_system_alist could have symbols instead of regexps; the
> > symbol would match the buffer's major mode.
> 
> I don't understand how that solve the problem.  When
> find-operation-coding-system is called, even if a buffer is
> given, the major mode of the buffer is not yet decided.

And on top of that, what useful information can we extract from a
major mode symbol wrt the correct decoding?  Some major modes imply
no-conversion, but that's about the only thing you can learn from a
major mode, since most major modes can go with _any_ encoding
whatsoever.  If we want to deduce no-conversion, why not say that
explicitly instead of going through a major mode?