On Mon, May 21, 2007 at 11:56:23PM +0200, Lars Marowsky-Bree wrote: > On 2007-05-21T17:51:20, Dejan Muhamedagic <[EMAIL PROTECTED]> wrote: > > > $ hg log -r cc868e409675 lrm/lrmd/lrmd.c > > $ hg log -r d2639cbc3351 lrm/lrmd/lrmd.c > > > > show nothing. Does that mean that these changesets have nothing to > > do with the lrmd.c? > > That's because d2639cbc3351 doesn't change lrm/lrmd/lrmd.c _relative to > it's first parent_ - but it would have shown some relative to the second > parent. It's a merge, not a classic diff. > > See my other mail for the longer explanation and a script to trace > this. > > > Alan's changes were in my local repository and I never did > > anything special to preserve them. Actually, it was Alan who > > noticed that his patch became part of mine while he has been > > reviewing it. The funny thing is that they were there all the time > > and I tend to pull changes fairly regularly. > > This doesn't seem to be perfectly consistent with the history in the > repository. I'd guess that at some point, possibly at the one where you > blew away your repo on May 7th or later, you had changes to lrmd.c which > weren't committed, and which you pulled over to the new repo - and as > part of that, you also pulled over Alan's changes. > > Your mail about selectively committing apache.in and being notified that > trying to push would create remote branches suggests something like > that - maybe you had to commit before the merge or something and ended > up with a branch - because you can't merge with outstanding uncommitted > branches. I might even guess somewhere, somewhen, -f was involved. ;-)
There wasn't and I think that I told you what I did: create a new repository and copy files using cp(1) from the old one and then do commit/push. I'm so scared of this thing, I won't use the -f in a thousand years :) > I think both Alan and you use Mercurial to frequently only commit "one" > file while you have several local changes, because in this way, you can > work on "independent" changes w/o actually creating several work-spaces. Indeed. I almost always do hg commit file1 file2 ... > When you then commit and merge (selectively as well), seemingly weird > things sometimes happen to you. > > > I think > http://www.selenic.com/mercurial/wiki/index.cgi/MultipleCommitters is a > good place for some more hints about this. Thanks. I'll take a look. > > > Regards, > Lars > > -- > Teamlead Kernel, SuSE Labs, Research and Development > SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG N?rnberg) > "Experience is the name everyone gives to their mistakes." -- Oscar Wilde > > _______________________________________________________ > Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev > Home Page: http://linux-ha.org/ -- Dejan _______________________________________________________ Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev Home Page: http://linux-ha.org/
