Bundling tools together with hg for win32

Steve Borho steve at borho.org
Tue Aug 21 18:41:50 CDT 2007


On Tuesday 21 August 2007 04:12:01 pm Steve Borho wrote:
> On Tuesday 21 August 2007 3:04:29 pm Patrick Mézard wrote:
> > Steve Borho a écrit :
> > > On Wednesday 15 August 2007 5:00:04 pm Steve Borho wrote:
> > > Unfortunately I'm currently hung up on making MQ work on win32 with
> > > the tip of crew or main.  There were some patches applied recently
> > > to make gnuwin32-patch work correctly (by passing --binary), but
> > > those patches dissapeared when the mpatch work was rolled in, and
> > > it's not obvious to me how to make mpatch work correctly.  I
> > > haven't found any documentation about how the patch code on the tip
> > > is expected to work.
> >
> > The old code is still in crew, but not thoroughly tested. To activate
> > it, set the ui.patch configuration option to the patch command to
> > call like:
> >
> > [ui]
> > patch = patch
> >
> > The --binary options should still be passed under Windows.
>
> So if I don't specify a patch tool in [ui] it will use mpatch by
> default?  Part of the problem I have is that I can't tell which
> implementation it's using.
>
> > mpatch is supposed to work out of the box. It evaluates its input in
> > binary mode so you may have issues while trying to apply an LF patch
> > to a CRLF source or the opposite. Otherwise, it should work, I use it
> > every day with Windows CRLF sources or with mercurial LF codebase
> > (using recent crew builds).
>
> Interesting.  It seems to work correctly if I separate the CRLF patches
> into their own patch.  It's when I combine LF and CRLF changes together
> that the qpush operation fails.

<self-responding twice in one day, that's got to be some kind of record>

I hit send too soon.  I haven't been able to reproduce the initial problem 
since I removed the patch specification in my config files and unfortunately 
I deleted the patch which was refusing to apply, so I'm at a loss at how to 
make it happen again.  Please disregard this whole mess.

I'll try to post something to the list when at least a few of the goals in the 
checklist are completed.

-- 
Steve Borho (steve at borho.org)
http://www.borho.org/~steve/steve.asc
Key fingerprint = 2D08 E7CF B624 624C DE1F  E2E4 B0C2 5292 F2C6 2C8C



More information about the Mercurial-devel mailing list