Le mardi 23 mars 2010 10:42:00, Senthil Kumaran a écrit :

I had to reinstall my OS recently and I checked out pristine copies of trunk,py3k,release26-maint,release31-maint.

While svnmerge.py init on release26-maint and release31-maint worked fine and I was able to merge a revision to release26-maint.

On py3k, the svn init gave the following problem.

Why are you using svn init? It was already done in all branches.

-- Victor Stinner http://www.haypocalc.com/

Le mardi 23 mars 2010 12:00:15, vous avez écrit :

On Tue, Mar 23, 2010 at 3:57 PM, Victor Stinner

victor.stinner@haypocalc.com wrote:

...

Why are you using svn init? It was already done in all branches.

I had this understanding that, while using svnmerge.py you might to do one-time operation on svnmerge.py init on local repo and then do the svnmerge.py merge -r xxx http://www.python.org/dev/faq/#how-do-i-prepare-a-new-branch-for-merging

"You need to initialize a *new branch* ...": release26-maint and release31- maint are not new branches ;-)

Now that i have done it release26-maint and release31-maint, does it create any problem?

svnmerge.py uses two 2 properties (svnmerge-blocked and svnmerge-integrated) on the directory.

Let's try on a fresh release26-maint checkout:

$ svn pl . Propriétés sur '.' svn:ignore svnmerge-blocked svnmerge-integrated svk:merge

$ svn pg svnmerge-integrated . /python/trunk:1-66720,66723-66743,66746-66751,66753-66755,...,79311,79325,79329

$ svn pg svnmerge-blocked . /python/trunk:66721-66722,66744-66745,...,79208,79294

It's look like the repository is ok. If you loose and cleared these properties in your local checkout, use "svn revert ." to restore the properties.

-- Victor Stinner http://www.haypocalc.com/

On Wed, 24 Mar 2010 01:13:54 +1000, Nick Coghlan ncoghlan@gmail.com wrote:

I'm not sure if there should be any change in the wording of the FAQ entry for this - maybe expanding on the last sentence a bit?

Current: "This is a one-time operation." Possible change: "This is a one-time operation (i.e. only when the branch is originally created, not when each developer creates a local checkout for the branch)."

+1. I was bitten by this when I first started, and I think your proposed wording explains it well and would have kept me from making a mistake.

-- R. David Murray www.bitdance.com

Alexandre Vassalotti wrote:

On Tue, Mar 23, 2010 at 11:13 AM, Nick Coghlan ncoghlan@gmail.com wrote:

...

I'm not sure if there should be any change in the wording of the FAQ entry for this - maybe expanding on the last sentence a bit?

Could we put a commit hook to prevent committing unintentional "svnmerge.py init"?

At this point it's probably not worth putting more effort into svn infrastructure.

-- Eric.

On Tue, Mar 23, 2010 at 08:13, Nick Coghlan ncoghlan@gmail.com wrote:

Victor Stinner wrote:

...

It's look like the repository is ok.

I confirmed this by looking at what Senthil's commit actually did on the Python 2.6 branch - it just added some redundant info to the beginning of the svnmerge-integrated property.

As that redundant info is no longer present, I expect svnmerge cleared it out on a subsequent merge command.

I'm not sure if there should be any change in the wording of the FAQ entry for this - maybe expanding on the last sentence a bit?

Current: "This is a one-time operation." Possible change: "This is a one-time operation (i.e. only when the branch is originally created, not when each developer creates a local checkout for the branch)."

Done.

-Brett

Cheers, Nick.

-- Nick Coghlan | ncoghlan@gmail.com | Brisbane, Australia

---

python-committers mailing list python-committers@python.org http://mail.python.org/mailman/listinfo/python-committers