On Mon, Aug 16, 2010 at 3:40 AM, Julian Foad <julian.foad_at_wandisco.com> wrote:
> On Thu, 2010-08-12, Paul Burba wrote:
>> I'm working on that approach, but in the meantime I have a question
>> about the behavior of svn diff --summarize when describing the
>> addition of paths with properties.
>> ### Let's take a vanilla greek tree, add a file and directory,
>> ### set a prop on each, and commit as r2:
>> >echo nu file > nu
>> >svn add nu
>> A nu
>> >svn ps prop val nu
>> property 'prop' set on 'nu'
>> >svn mkdir J
>> A J
>> >svn ps prop val J
>> property 'prop' set on 'J'
>> ### Status looks right, two scheduled additions:
>> >svn st
>> A nu
>> A J
>> ### Commit it
>> >svn ci -m ""
>> Adding J
>> Adding nu
>> Transmitting file data .
>> Committed revision 2.
>> ### Check diff --summarize for r2
>> >svn diff -r1:2 --summarize
>> AM nu
>> AM J
>> Should we really be describing the props for this added (not copied!)
>> file as modified? It seems incorrect to call the props modified in
>> this case. I ask because this is one of the problems I am running
>> into with fixing this issue in mod_dav_svn, and before I spend more
>> time trying to "fix" it I want to be sure the current behavior is
>> really what we expect.
>> Paul "Dying slowly at the hands of mod_dav_svn" Burba
> I would suggest that the output of "svn diff --summarize" should be
> consistent with the likes of "svn add" (no 'M'), "svn status" (no 'M')
> and "svn update" (I haven't checked this one).
(I mentioned this in response to the commit, but also will do so here
The other place we're inconsistent is the template shown in the editor
used for log messages when 'svn ci' is run without -F or -m. I
believe it still shows 'AM'. (In fact, this was one of the earliest
bugs I attempted to fix when I first started hacking Subversion
several years ago. Ah, nostalgia.)
Received on 2010-08-16 21:35:41 CEST