Re: Vowel with Umlaut

[David Kastrup]

Nick Payne <address@hidden> writes:

> On 19/10/11 05:09, GRAEME F ST CLAIR wrote:
>> General reply to Messrs Rogers, Peekay and Kastrup!
>>
>> In the end I googled lilypond and found the \char approach myself.
>> Right now, I'm settling for it, because as DK hints, both TextPad
>> and jEdit will save UTF-8 just fine, but the next time you open the
>> file, you see mysterious blobs, not the intended character.  That's
>> why I committed "mangling" on my previous project!
>>
>> But I intend to re-visit some German songs I lily'ed a few years
>> ago, and obviously \char won't really hack it - I like the
>> suggestion of Emacs, and will go looking for it.
>
> jEdit works fine for me on both Linux and Windows in preserving
> extended characters across editing sessions. Just open Global Options
> from the Utilities menu and make sure that under encodings the default
> character encoding is UTF-8.

Most editors should preserve text when you edit in the correct encoding.
The advantage of Emacs is that it is rather good at preserving the
_byte_ stream of unedited text even when in the wrong encoding (Emacs
actually also is rather good at detecting the coding of a file, so you
are not all that likely to even start creating mish-mash files even if
you don't realise that somebody sent you something differently encoded
from what you expect).  So if you edit a file as utf-8 and have some
latin-1 passages in it originally, the latin-1 passages will still be
the same when you open the file as latin-1, or finally discover them and
use M-x recode-region RET on them, telling Emacs that the characters
were really latin-1 and wrongly interpreted as utf-8.

Other editors turn latin-1 passages into increasingly unrecoverable crap
each time you load and save under utf-8.  The wrongness of the encoding
_deteriorates_.  If your file is encoded inconsistently in different
parts, Emacs can't magically fix this, but it won't make the situation
worse.  And that is quite comforting.

-- 
David Kastrup