|
|
lars hofhansl 2012-08-24, 17:41
0.94.1 has been released two weeks ago. Seeing that it typically takes multiple weeks to tie up a release go through the RC process, in the interest of a faster release cycle, I would like to start thinking about 0.94.2 now. There are 38 improvements and bugfixes against 0.94.2 already, and some of them include important performance fixes.
Thoughts? Keep up the pace? Or slow down?
Thanks. -- Lars
On Fri, Aug 24, 2012 at 10:41 AM, lars hofhansl <[EMAIL PROTECTED]> wrote: > 0.94.1 has been released two weeks ago. Seeing that it typically takes multiple weeks to tie up a release go through the RC process, in the interest of a faster release cycle, > I would like to start thinking about 0.94.2 now. > There are 38 improvements and bugfixes against 0.94.2 already, and some of them include important performance fixes. > > Thoughts? Keep up the pace? Or slow down? >
I'd be up for reviewing an RC. 38 changes easily justifies a point release and there are a few corner case regressions whose fixes should be out sooner rather than later.
St.Ack
Elliott Clark 2012-08-24, 18:23
I think the perf improvements alone would warrant a 0.94.2. HBASE-6602 was pretty critical for us and HBASE-6603 was a great find.
Smaller releases mean more work testing rc's but it sure is nice to debug when the delta is smaller (see HBASE-6577).
On Fri, Aug 24, 2012 at 11:17 AM, Stack <[EMAIL PROTECTED]> wrote: > On Fri, Aug 24, 2012 at 10:41 AM, lars hofhansl <[EMAIL PROTECTED]> wrote: >> 0.94.1 has been released two weeks ago. Seeing that it typically takes multiple weeks to tie up a release go through the RC process, in the interest of a faster release cycle, >> I would like to start thinking about 0.94.2 now. >> There are 38 improvements and bugfixes against 0.94.2 already, and some of them include important performance fixes. >> >> Thoughts? Keep up the pace? Or slow down? >> > > I'd be up for reviewing an RC. 38 changes easily justifies a point > release and there are a few corner case regressions whose fixes should > be out sooner rather than later. > > St.Ack
lars hofhansl 2012-08-24, 19:06
Alright. I created a 0.94.3 release target. If you can, please move all 0.94.x jiras you don't plan to work on soon (within a week or so) to 0.94.3. I'll do a sweep throughout the next week. Maybe I can cut an RC by the end of next week. Thanks.
-- Lars
----- Original Message ----- From: Elliott Clark <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Cc: lars hofhansl <[EMAIL PROTECTED]> Sent: Friday, August 24, 2012 11:23 AM Subject: Re: 0.94.2
I think the perf improvements alone would warrant a 0.94.2. HBASE-6602 was pretty critical for us and HBASE-6603 was a great find.
Smaller releases mean more work testing rc's but it sure is nice to debug when the delta is smaller (see HBASE-6577).
On Fri, Aug 24, 2012 at 11:17 AM, Stack <[EMAIL PROTECTED]> wrote: > On Fri, Aug 24, 2012 at 10:41 AM, lars hofhansl <[EMAIL PROTECTED]> wrote: >> 0.94.1 has been released two weeks ago. Seeing that it typically takes multiple weeks to tie up a release go through the RC process, in the interest of a faster release cycle, >> I would like to start thinking about 0.94.2 now. >> There are 38 improvements and bugfixes against 0.94.2 already, and some of them include important performance fixes. >> >> Thoughts? Keep up the pace? Or slow down? >> > > I'd be up for reviewing an RC. 38 changes easily justifies a point > release and there are a few corner case regressions whose fixes should > be out sooner rather than later. > > St.Ack
|
|