[PATCH] windows: switch the console encoding into an encoding actually used (issue2926)

[Matt Mackall]

On Fri, 2011-07-29 at 17:25 +0200, Adrian Buehlmann wrote:
> On 2011-07-29 17:20, Matt Mackall wrote:
> > On Fri, 2011-07-29 at 17:00 +0200, Adrian Buehlmann wrote:
> >> On 2011-07-29 02:20, Andrei Polushin wrote:
> >>> 29.07.2011 7:00, Adrian Buehlmann wrote:
> >>>> On 2011-07-29 01:26, Andrei Polushin wrote:
> >>>>> Probably, I need help to do it in a non-natural way. Is it worth the effort?
> >>>>> Looks like Mercurial already has plaform-specific C code anyway.
> >>>>>
> >>>>> I plan to leave C code in osutil.c and call it from the windows-specific
> >>>>> extension. Could it be done this way?
> >>>>
> >>>> It could, but for newer things we prefer not adding C code just to call
> >>>> a few Windows API's. Mercurial on Windows now depends on ctypes.
> >>>>
> >>>> I could help you by adding basic infrastructure in win32.py. It's not
> >>>> that difficult. You could then call that in your extension.
> >>>
> >>> Well, I'm not sure. My concern is that the problem with redirection is not
> >>> finally solved.
> > 
> > Adrian, did you read this part?
> > 
> > This code is still a proof-of-concept.
> > 
> 
> Yes I did.
> 
> Does that mean you'll apply the C code changes?

Why would I apply a proof of concept?

I think your objection is premature - this code is interesting but not
ready for inclusion. And I think it's not improbable that Andrei will
discover that doing it right with pipes is either impossible or more
involved than ctypes can reasonably handle.

-- 
Mathematics is the supreme nostalgia of our time.