Re: mule: why not show the damage of choosing "raw-text" etc. now?

[Eli Zaretskii <eliz@is.elta.co.il>]

On 24 Apr 2002, Dan Jacobson wrote:

> I do a certain M-x compile command that produces output that looks
> like normal big5 chinese... at least the first page of which that I
> see in the split window.
> 
> I attempt to write this to disk with C-x C-w some_filename.
> 
> Due to some non-big5 character somewhere deep in the output, I encounter
> 
>    These default coding systems were tried:
>      chinese-big5-unix
>    However, none of them safely encodes the target text.

Does it help to say "C-x RET c big5 RET M-x compile RET" instead?

> My point is, _why allow the file to keep on looking good this
> session?_  If you are going to turn it into a "unrecoverable bunch of
> <98>'s", why not do it now in front of the user's face instead of
> having him think that the file is A-OK, and can be e-mailed to
> friends, or is ready to go for the big presentation tomorrow?

Because the Mule design is that Emacs never considers how it will encode 
the file until such time as you actually ask it to do so.  One problem 
with what you suggest is that yanking some text that cannot be encoded in 
the buffer's coding system will trigger annoying questions or even 
display gibberish, even though all you want is remove some of the text 
you've yanked, or maybe save the buffer with a different encoding.