While merging branches for plip 9214 I ran into a couple of problems,
which some of you may run into as well, so here are my notes.  When
doing a simple "svn merge
http://.../Plone/branches/plip9214-emaillogins"; into the Plone 4.0
branch, several changes were pulled in that definitely were not done
on my branch.  I do not know out of which hat subversion (1.5.4)
pulled these.  I had to solve some conflicts and retry the merge a
couple of times.  It resulted in the following changed, added or
removed files that I did not actually touch in my branch:

M      Products/CMFPlone/profiles/default/cssregistry.xml
M      Products/CMFPlone/profiles/default/propertiestool.xml
M      Products/CMFPlone/profiles/default/actions.xml
A  +   Products/CMFPlone/skins/plone_styles/fullscreen.css.dtml
A  +   Products/CMFPlone/skins/plone_styles/fullscreen.css.dtml.metadata
M      Products/CMFPlone/FactoryTool.py
M      Products/CMFPlone/tests/testNavTree.py
M      Products/CMFPlone/tests/testInterfaces.py
D      Products/CMFPlone/tests/testPloneFolder.py
M      Products/CMFPlone/tests/testPortalCreation.py
M      Products/CMFPlone/PloneTool.py
M      Products/CMFPlone/deprecated.zcml

An attempt at first merging the 4.0 branch to my branch gave the same
kinds of problems.

The solution is easy enough.  Do not use the newly sprouted automatic
merge capabilities of subversion, but specify the revisions by hand.
In my case:

svn merge -r28396:HEAD \

Of course if an automatic merge works for you, that is fine.  But be
warned, and double check that you do not miss any changes and do not
introduce unwanted changes.

For a few other packages the automatic merge failed with a different

svn merge ..../plone.app.upgrade/branches/plip9214-emaillogins
svn: Target path does not exist

Again, specifying a specific revision range (or in this case one
specific changeset) helped.


Maurits van Rees | http://maurits.vanrees.org/
            Work | http://zestsoftware.nl/
"Yes, we study economics,
got to make the money make some kind of sense."  - Geoff Mann

Framework-Team mailing list

Reply via email to