A proposal on solve encoding problem on Windows.

[Roger Gammans]

On Fri, Oct 21, 2011 at 03:13:13AM -0700, Andrey wrote:
> 
> 
> > Instead of running a mode guess algorithm, couldn't mercurial
> > store a repo encoding field. Which if blank /non-exsitant  mercurial
> > defaults to the current behavior ?
> >
> Do you mean that the whole repository must be configured to contain only 
> Cp1252 or UTF-16 encoded names ?
> Do you mean to change the manifest file to keep the encoding information per 
> file ?

Where it is stored is an implementation detail. Except I believe it should
be revision controled.

But do you really have a use-case for files in a single repo *stored* 
with different encodings. It strikes me as that way madness lies.

I have to say I can't imagine one, doesnt utf8 encode all the same codepoints 
as cp-1252?

Have I missed something - does the windows wide character API do something odd
when given the top-portion of cp1252?

TTFN
-- 
Roger. 	                        Home| http://www.sandman.uklinux.net/
Master of Peng Shui.      (Ancient oriental art of Penguin Arranging)
Work|Independent Sys Consultant | http://www.computer-surgery.co.uk/
 New key Fpr: 0F2F E1DF 4CD2 5E7B EF9F  B173 4CFA F143 ADBE 6B00