SV: [PATCH 0 of 2] remove encodedir/decodedir from filelog
henrik.stuart at edlund.dk
Tue May 26 04:53:27 CDT 2009
The problem I noticed is not the same as brendan has, and I believe the problem I alluded to has been there for a long while.
Neither Benoit, nor I, have been able to reproduce brendan's problem, and as such, we await his return.
> -----Oprindelig meddelelse-----
> Fra: mercurial-devel-bounces at selenic.com [mailto:mercurial-devel-
> bounces at selenic.com] På vegne af Adrian Buehlmann
> Sendt: 26. maj 2009 11:24
> Til: Benoit Boissinot
> Cc: mercurial-devel at selenic.com
> Emne: Re: [PATCH 0 of 2] remove encodedir/decodedir from filelog
> On 20.05.2009 16:06, Benoit Boissinot wrote:
> > This patch serie removes the layering violation due to
> > the .i/.d escaping done in filelog.
> > 1) refactor the fncache to make 2) easier
> > 2) move the encodedir/decodedir in store, takes care of
> > backward compat (streamclone, fncache on-disk format)
> (this is now in crew as 03196ac9a8b9 and 810387f59696)
> Some late comment:
> This seems to have been introduced some new problems with
> the dirstate as brandan and hstuart have noted on #mercurial.
> I'll post a message of hstuart from irc that may be
> helpful in testing:
> < hstuart> abuehl: have a common base commit then in one commit
> make a directory with name foo.i then up to the common base then
> make a file with name foo.i and commit then try to merge :)
> Trying to "fix" layering violations is nice, but we should
> be careful not to introduce new problems with risky
> Don't fix it if it ain't broken ;)
> Mercurial-devel mailing list
> Mercurial-devel at selenic.com
More information about the Mercurial-devel