db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Bernt M. Johnsen" <Bernt.John...@Sun.COM>
Subject Re: Derby commit process
Date Tue, 28 Feb 2006 21:49:12 GMT
>>>>>>>>>>>> Kathey Marsden wrote (2006-02-23 09:25:03):
> Bryan Pendleton wrote:
> 
> > In preparation for issuing my first commit, I tried to write up an
> > overall
> > summary of the Derby commit process as I understand it. Could you please
> > give it a quick read and help me improve it?
> >
> > http://wiki.apache.org/db-derby/DerbyCommitProcess
> >
> Thanks Bryan for writing this up. I only have a few comments, mostly 
> about things I have done differently in the past which may or may not be
> correct:
> 
> 1) " The committer applies the patch to the current trunk and verifies
> that it applies cleanly and passes derbyall."
> 
> I do not run derbyall for every patch, having assumed that the
> contributor did that.  I do run specific relevant tests and will usually
> run derbynetmats for network server changes as it is not included in
> derbyall if folks don't download JCC and sometimes J2ME/CDC/Foundation 
> or jdk131 if I think there will be impact..

I think that there should be a definition of a "minimum acceptance
test" which all committers should run. Personally I prefer to run
derbyall because I don't trust that I'm (not yet) able to infer all
possible consequences of a patch (it has been a problem lately since
derbyall has been kind of unreliable and that a lot of time hase been
used to figure out that the failures was not relevant to the patch).

> 
> <>2)" The committer updates the JIRA entry to:
>     -  mark the issue as resolved"
> 
> I have always left this for the fixer to do, but maybe it makes more
> sense for the committer to do it.  I don't have any objection.  Maybe
> modify this to say:
> - mark the issue as resolved if it has been fully resolved with the
> patch.

I think the developer/fixer should do this. Ifv the committer should
do it, he/she would very often have to ask the fixer if the patch is
partial vs. the issue or the issoe should be resolved.

> 
> 3) There is some special svn processing needed for added/deleted files
> in patches.  Perhaps some explanation of that would be good or a link to
> someplace that does. 

Yes.

> 4) I seem to remember Oyvind (Oyvind.Bakksjo@Sun.COM) had  posted some
> tips for committers/reviewers at one point , but I can't seem to find
> them in the archives now.

I'll see if I can find something.

> 5) Would a link to http://incubator.apache.org/guides/committer.html be
> appropriate?

Yes.


-- 
Bernt Marius Johnsen, Database Technology Group, 
Staff Engineer, Technical Lead Derby/Java DB
Sun Microsystems, Trondheim, Norway

Mime
View raw message