Mule bugs: misidentification (Latin-1 vs. Chinese), revert issues

[Nix]

On 27 Oct 2006, stephen at xemacs.org verbalised:
> Actually, you can get quite far with display without MULE.  IIRC,
> historically that's why Gnus does things this way.  Consider that X11
> doesn't know any differences among ISO Latin-1, ISO Latin-9 (hardly
> surprising to humans), and ISO Arabic (astonishing to most humans).
> Just change the registry of the font for the default face from
> iso8859-1 to iso8859-15 to iso8859-6 as you have need for the ancient
> currency symbol (aka the "horny bullet"), the Euro, or the language of
> the Koran.

Well, you have to be able to determine what fonts to use, i.e., what
encoding things are likely in. (But file-coding mode can do that on its
own, at least for a subset of encodings.)

> It's multilingual input and handling files in different character sets
> *simultaneously* where MULE (or something like it) becomes essential.

Oh yes. Of course anyone who speaks several not-closely-related
languages eventually has to deal with *that* snake-pit, and except for
the Emacsen most software is *bad* at it. (But you'd know that much
better than I.)

>  > OK. I can provide my 14 local patches, too (some already present in 21.4
>  > but I didn't port them forward because I'm a moron).
>
> Day job etc keeping you busy, eh?  :-)  Yes, please do.

Coming up (well, probably fewer than 14: I can remember that a few were
rejected and won't re-post those.)

-- 
`When we are born we have plenty of Hydrogen but as we age our
 Hydrogen pool becomes depleted.'