Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 9 Oct 2006 21:37:38 -0400
From:      John Baldwin <jhb@freebsd.org>
To:        John Birrell <jb@freebsd.org>
Cc:        Perforce Change Reviews <perforce@freebsd.org>
Subject:   Re: PERFORCE change 107544 for review
Message-ID:  <200610092137.39154.jhb@freebsd.org>
In-Reply-To: <200610091034.k99AYtcC064485@repoman.freebsd.org>
References:  <200610091034.k99AYtcC064485@repoman.freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Monday 09 October 2006 06:34, John Birrell wrote:
> http://perforce.freebsd.org/chv.cgi?CH=107544
> 
> Change 107544 by jb@jb_freebsd7 on 2006/10/09 10:21:52
> 
> 	This submit is the result of an integrate -ifb from current because it
> 	appears that p4 gets confused by integrating files from different branches
> 	which share the same file names. It can't or won't detect changes in the
> 	content of the files because of it's secret dea of what a revision is.

Read the docs. :)  When you do p4 integ -i, it assumes that your version of the
file is the master copy, so it leaves your version unchanged when a file is added
on the parent but already exists on the child branch and you do p4 integ -i.  Note
that because CVS commits are just submitted as new submits to //depot/vendor, p4
has no way to know that the import it gets from a CVS commit is really adding the
same file that was added over in a p4 branch.  If HEAD were actually in p4 and
and the original commit had been done via a reverse integrate up into HEAD from
the sun4v branch, p4 would be able to DTRT.  However, since HEAD is in CVS and we
clone a copy into p4, p4 doesn't have any history to know that a new file in
//depot/vendor/freebsd/... is really a commit of a file in a child branch.  This
is one reason why having the work trees and the actual HEAD in the same VCS would
be helpful.

-- 
John Baldwin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200610092137.39154.jhb>