lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Steven A Rowe <sar...@syr.edu>
Subject RE: switch jars to ivy mechanism?
Date Tue, 27 Mar 2012 20:50:03 GMT
On 3/27/2012 at 4:38 PM, Robert Muir wrote:
> On Tue, Mar 27, 2012 at 4:33 PM, Steven A Rowe <sarowe@syr.edu> wrote:
> > Robert,
> >
> > I disagree with you: I think Maven really is an option.  However,
> > I do agree that it would be significantly more work, and I recognize
> > that lots of devs here loathe Maven, so I will not tilt against this
> > particular windmill.
> >
> > On the technical arguments, though: a) there *is* a JFlex maven 
> > plugin; and b) if you follow the link you gave to the Jenkins Maven 
> > trunk build script, you'll see that the one test which was being 
> > ignored is no longer ignored -- that chunk of the shell script is 
> > commmented out -- Maven currently runs all tests and passes them 
> > just as often as the Ant builds.  (BasicDistributedZkTest is now an
> > unhappy camper no matter which camp it's in these days.)
>
> Maven refers to the maven build we have: it doesn't support all of
> our features. This isn't really a debate, its a fact. I don't argue
> that theoretically it can't in the future, but it does not do so
> right now. There are a lot of missing tasks (jflex generation was
> just a simple one, but there are lots of little things like this).
> Sure, it might be possible they could be added, but thats a ton more
> work than just addressing jar dependencies.

Agreed.  And also besides plus additionally, Lucene/Solr has way more Ant hackers than Maven
hackers.

Steve

Mime
View raw message