mahout-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sean Owen <sro...@gmail.com>
Subject Re: Assign hack slowdown
Date Tue, 02 Mar 2010 16:56:09 GMT
Yeah that's good. No I don't want to be heavy-handed about it, but I
also know there will never otherwise be a time when there aren't
issues that could go in real soon. With any luck 0.20.2 lands in Maven
shortly, I create the release and we ship it in a couple days and
resume normal life. Yeah, we should aim for quick release periods so
they're routine and non-disruptive. This 0.20.2 thing is really
dragging on, but the quiet period of submitting largely polish and
improvements has been good.

Use your judgment as in all things. If you check it in tomorrow and it
goes out Friday -- how confident are you there's not a bug, that sort
of thing, since this will live as the official release for months.



On Tue, Mar 2, 2010 at 4:45 PM, Jake Mannix <jake.mannix@gmail.com> wrote:
> So it looks like we've actually been more productive during this "code
> freeze" than the previous few weeks before it.
>
> You saying we get one last day (today), to put in patches which are
> sitting on JIRA tickets ready to go in (I've tried to be good, and only
> post patches and not check in, as hard as it's been :) ) ?
>
> I made some significant improvements in DistributedRowMatrix over
> the past week (most in the latest patch for MAHOUT-310), which
> would be nice to put in (they have unit tests, and are backwards
> compatible).
>
>  -jake
>
>
>
>
>> On Tue, Mar 2, 2010 at 9:45 AM, Robin Anil <robin.anil@gmail.com> wrote:
>> > Another issue along the same topic
>> >
>> > We have Bigram in co-occurrence, IntPairWritable in LDA, and gramkey in
>> > dictionary vectorizer ? they store the same thing two integers. Can't we
>> > start a crackdown on these and others like these?
>> >
>> > Robin
>> >
>>
>

Mime
View raw message