How this about this: post-migration, I'll annotate README (the one in our
top-level directory) with all this information, plus the final record of the
revision offset caused by the load into the ASF repository. (And if I can
dig up the relevant info, I'll do a post-facto documentation of the merge of
the svn-logos/ repository data long ago, too.)
Greg Stein wrote:
> Awesome. Could you drop the text below into a file somewhere, for
> future archeologists?
>
> On Sat, Nov 14, 2009 at 22:08, C. Michael Pilato <cmpilato_at_collab.net> wrote:
>> Karl Fogel wrote:
>>> Blair Zajac <blair_at_orcaware.com> writes:
>>>> I think that's what Michael did.
>>>>
>>>> https://issues.apache.org/jira/browse/INFRA-2321?focusedCommentId=12776647&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#action_12776647
>>> Fascinating. So he did. The history's a bit wonky around the time of
>>> the switchover. For example, this was r2:
>>>
>>> http://svn.collab.net/viewvc/svn-complete?view=revision&revision=3657
>>>
>>> But http://svn.collab.net/viewvc/svn-complete?view=revision&revision=3656
>>> is not quite the same as r1. No matter; it's still cool :-).
>> The CVS repository had some trailing revisions which occurred after the
>> switchover. Most of them were commits to www/ (which we manually mirrored
>> based on our subversion commits to trunk/www/). A couple of them were
>> things like system-wide automated tweaks to www/robots.txt made by CollabNet
>> folk. Also, we had real tags and branches in our CVS repos that we didn't
>> bring with us into Subversion. So what I did was this:
>>
>> 1. Converted all the CVS history to Subversion (into repos 'svn-from-cvs')
>>
>> 2. Dumped all the revisions from 'svn-from-cvs' up to the switchover time
>> (r1:3654) -- *not* the extra/duplicate revs mentioned above -- and loaded
>> them into a new repos, 'svn-complete'.
>>
>> 3. Committed the deletion of all our tags and branches, leaving the
>> directories themselves, from 'svn-complete'.
>>
>> 4. Since r1 in the 'svn' repos was a massive import, that would have
>> clashed mightily being loaded atop already-existing files and directories in
>> 'svn-complete'. So I checked out svn-complete/trunk_at_HEAD, then exported
>> svn/trunk_at_r1 atop it. The local mods were the small delta between what we
>> got outta CVS on August 31, 2001, and what we put into Subversion. They
>> were mostly the result of $Date$ keyword formatting differences. I
>> committed those local mods, which now brought 'svn-complete' in sync with
>> 'svn_at_r1' *except* that 'svn-complete' still had empty tags and branches
>> directories (whereas we didn't create those in the 'svn' repository until
>> later when we needed them).
>>
>> 5. Dumped -r2:531 of 'svn', loading into 'svn-complete'.
>>
>> 6. Made r532 from 'svn' a no-op in 'svn-complete', since that was the
>> revision in which we created the tags/ directory.[1]
>>
>> 5. Dumped -r533:1236 of 'svn', loading into 'svn-complete'.
>>
>> 7. Made r1237 from 'svn' a no-op in 'svn-complete', since that was the
>> revision in which we created the branches/ directory.[2]
>>
>> 8. And then dumped the rest of 'svn', loading into 'svn-complete'.
>>
>> I'm actually pretty happy with the stitchwork.
>>
>> -- C-Mike
>>
>> [1] http://svn.collab.net/viewvc/svn-complete?view=revision&revision=4187
>> [2] http://svn.collab.net/viewvc/svn-complete?view=revision&revision=4892
>>
>> --
>> C. Michael Pilato <cmpilato_at_collab.net>
>> CollabNet <> www.collab.net <> Distributed Development On Demand
>>
>> ------------------------------------------------------
>> http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2418059
--
C. Michael Pilato <cmpilato_at_collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2418070
Received on 2009-11-15 06:31:26 CET