Getting 2.4 Released

[John Ralls]

On Oct 16, 2010, at 12:57 PM, John Ralls wrote:

> 
> On Oct 16, 2010, at 10:47 AM, jh wrote:
> 
>> Hi,
>> On Sat, 2010-10-16 at 10:06 -0700, John Ralls wrote:
>> ...
>>> What needs to be done to make the next test release a "release candidate" that we can safely recommend to ordinary users that they can try it out without risking their data? (FWIW, I've been using 2.3.15 regularly without incident, but there are lots of features that I don't use.)
>>> 
>> I still get a couple of failures (with r19666) and errors, when I run
>> 'make check'. Doesn't that need to be fixed?
>> 
>> BTW, how can I get a nice list of what goes wrong?
>> when I enter 
>> 
>> make check > check.log
>> 
>> the output seems to be devided between the screen and the log file.
>> 
> 
> To get errors into the log, use
> 
> make check > check.log 2>&1
> 
> Well, I think that make check failures should be fixed before a release, but I don't think that I'm the "decider". But do please write a bug ticket against the check failures if there isn't one already.
> 
> Please remember to copy the list on your replies ("reply all" or "reply list" on most mail clients).
> 


Having said that, I found the failures on OSX were caused by a couple of dumb mistakes in the test Makefiles. I've already checked in the fix for backend/dbi, and I'm checking on Linux now to make sure that my correction in business/business-core doesn't break anything there. (It shouldn't.)

There are a bunch of errors in the Python bindings (on OSX), but those aren't fully supported at the moment and I wouldn't recommend holding up the release for them.

Regards,
John Ralls