[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: svn merge --accept HUH ?

From: Paul Hammant <paul_at_hammant.org>
Date: Thu, 17 Sep 2009 14:02:05 -0500


What I'm reporting is that use of --diff3-cmd cancels any use of --
In this thread I'm not so much worried about our still breaking merge.

I would wish that --diff3-cmd is ignored for cases where people use --
accept, and not silently do something other than what you want. I'm
going to raise a Svn issue around that specifically :)


- Paul

On Sep 17, 2009, at 1:29 PM, Michael Diers wrote:

> Hash: RIPEMD160
> Paul Hammant wrote:
>> OK, found it ..
>> .. for --accept to be processed in any way whatsoever, /etc/
>> subversion/
>> config can't contain any entries for diff3
>> One presumes the same would be true on if you specified --diff3-cmd
>> on
>> the command line (which we were not).
> Still won't work in the general case, like in the presence of a tree
> conflict on directory hierarchy.
> Currently, --accept can only deal with what's already in the working
> copy. "Their" changes don't make it into the working copy in the
> case of
> tree-conflicted directories.
> I think 1.6.5 (or was it 1.6.6) has a patch that tells the user so.
> If you want to resolve the conflict as "theirs-full", you need to
> somehow copy that data into your working copy.
>> On Sep 11, 2009, at 10:37 AM, Paul Hammant wrote:
>>> I'm a seven year user of Svn, and new to this list, greetings folks.
>>> Before filing a bug, I'm to check here with y'all. Here's what I'm
>>> trying to do ...
>>> cd <my-trunk-dir>
>>> svn up
>>> svn merge --accept theirs-full <url-of-branch-to-merge-in> .
>>> The last list, despite the doco, offers merge conflicts for me to
>>> arbitrate over. We have kdiff3 linked in, but I suspect that the
>>> same
>>> would be true for the command line arbitration too. I expected to
>>> not
>>> be bothered by such issues given I specified "--accept theirs-full".
>>> It looks like (undocumented) that --accept=theirs-full is also
>>> acceptable for svn. Mis-spell any of those chars and svn will barf
>>> with un-recognised params.
>>> We're using Svn 1.6.3 on server and client. Day to day commits are
>>> working well, and we've merged before but without the --accept
>>> option.
>>> Have I got the syntax wrong? Or is this feature not working when it
>>> should be (for merge as opposed to resolve).
>>> Thoughts?
> - --
> Michael Diers, elego Software Solutions GmbH, http://www.elego.de
> Version: GnuPG v1.4.9 (Cygwin)
> iEYEAREDAAYFAkqygAkACgkQcEKlWnqVgz07zgCfUk8FxAZ3kzutFcVJDy7JCQTd
> Uc8An3nfw2gAehHogqhLq1TECsrG0qGK
> =Zr5m


To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_subversion.tigris.org].
Received on 2009-09-17 21:02:54 CEST

This is an archived mail posted to the Subversion Users mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.