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

Re: C-Mike's TODO for this week (was "Re: svn mergeinfo broken after merge of mergeinfo-less-copies branch")

From: C. Michael Pilato <cmpilato_at_collab.net>
Date: 2007-11-26 16:19:35 CET

David Glasser wrote:
> On Nov 24, 2007 8:56 PM, C. Michael Pilato <cmpilato@collab.net> wrote:
>> Mark Phippard wrote:
>>> I am updating a JavaHL test which looks at mergeinfo. It was of
>>> course failing because there is no longer mergeinfo from the copy. In
>>> the process of updating the test, I noticed a bug in the svn mergeinfo
>>> command. Consider this scenario:
>>>
>>> Standard Greek Tree
>>>
>>> r2: svn mkdir branches
>>> r3 svn cp A branches/A
>>> r4: changes and commit from A
>>> r5: merge to branches/A and commit
>>> r6: more changes and commit on A
>>>
>>> Now run:
>>>
>>> $ svn mergeinfo branches/a
>>> Path: branches/a
>>> Source path: /A
>>> Merged ranges: r2:4
>>> Eligible ranges: r1:2, r4:6
>>>
>>> The Eligible ranges should presumably not include r1:2 and just 4:6
>> 'svn mergeinfo' has not been updated to deal with the mergeinfoless-copies
>> changes. I'll be working on that this week [estimate: 8 hours], as well as:
>>
>> * Turning off mergeinfo calculations during 'svn merge' against pre-1.5
>> servers. [estimate: 4 hours]
>>
>> * Improving the default revisions chosen for a revision-less merge.
>> [estimate: 4 hours]
>>
>> * Changing the default path for a source-less copy to be "myself", which
>> is the only value with is both consistent and reasonably explainable
>> to the average user. [estimate: 1 hour]
>>
>> * Improving the svn_client_merge() API to actually do source-relatedness
>> checks instead of just assuming the sources aren't related. (And if
>> they *are* related, I'll probably just have it call
>> svn_client_merge_peg3() under the hood.) [estimate: 6 hours]
>
> Are you aware that many of the copy and merge tests are failing on the
> Mac buildbot over both DAV ra protocols since you merged your branch
> to trunk?

No sir, wasn't aware of that. Will look into that (on Linux only,
unfortunately) right now. (I tested only ra-local and ra-svn since my
changes didn't reach inside the RA layer.)

-- 
C. Michael Pilato <cmpilato@collab.net>
CollabNet   <>   www.collab.net   <>   Distributed Development On Demand

Received on Mon Nov 26 17:29:24 2007

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