On Feb 16, 2009, at 9:34 AM, Hyrum K. Wright wrote:
> On Feb 16, 2009, at 9:16 AM, Mark Phippard wrote:
>
>> It looks like all the blockers are complete and while the buildbot
>> has
>> been having some problems it looks like we are mostly green.
>>
>> I just ran the Serf tests and they all passed (that is one of the
>> buildbots that is red).
>>
>> The Windows buildbot is also red, the build is failing. I can
>> confirm
>> this. Here is the error:
>>
>> subversion\libsvn_fs_fs\rep-cache.c(35): error C2065:
>> 'REP_CACHE_DB_SQL' : undeclared identifier
>>
>> When I look in that file it is using a variable with that name, but
>> in
>> the rep-cache-db.h file that appears to be generated when you run
>> gen-make.py, I see a #define REP_CACHE_DB <-- NOTE no _SQL on the
>> end! I poked around the files and could not figure out where this is
>> being done.
>>
>> I had the same error on OSX until I re-ran autogen.sh so I am
>> guessing
>> some recent change was made which renamed this variable and the
>> Windows configuration was not updated?
>>
>> Anyway, presumably when the Windows build is fixed the tests will all
>> pass and we can proceed with branch and RC1?
>
> +1. As much as I hesitate to do it, I'd like to give people the
> better part of the morning (CST) to voice any final issues/concerns.
> So let's get the windows build issue fixed, and put the branch at 2000
> UTC (1400 CST / 1200 PST).
The windows build failure is fixed, so I continue to plan on branching
at 2000 UTC. Expect an RC by the end of the day.
We're still getting intermittent test failures in the ruby bindings,
but they aren't reproducible, so I don't think it should hold the
branch.
-Hyrum
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=1172478
Received on 2009-02-16 19:51:35 CET