On 04/10/2015 08:28 PM, Phil Steitz wrote: > On 4/9/15 10:08 AM, Thomas Neidhart wrote: >> On 04/09/2015 05:50 PM, Gilles wrote: >>> On Thu, 09 Apr 2015 17:44:08 +0200, Thomas Neidhart wrote: >>>> On 04/09/2015 11:20 AM, Luc Maisonobe wrote: >>>>> Hi all, >>>>> >>>>> Trying to get back on work, after a while ... >>>>> >>>>> I have just fixed a very annoying (and difficult to reproduce) bug >>>>> concerning BSP trees in 3.4.1. As I am on it, I will also try to >>>>> fix MATH-1211 created recently. >>>>> >>>>> The bug I fixed is really important as it breaks some other tools, >>>>> so I would need to release the fix officially as soon as possible. >>>>> >>>>> Therefore, I would like to publish a version 3.5 in the next few days. >>>>> Of course, I will do the release management hurdle. What do you >>>>> think about it? >>>> I see only one actual feature being added to 3.5 so far (the additional >>>> method introduced to Descriptive and SummaryStatistics). >>>> >>>> We could copy the current 3_X branch to a 3_4_X branch and remove this >>>> specific feature to really release a pure bugfix release for the 3.4 >>>> branch. >>> Alternatively, we could leave it (and I should add another similar before >>> release)... >>> Is there a practical reason that needs preserving the 3.5 option? >> not really imho, but it would be quite unusual for math to release >> another minor release so quickly (and almost entirely consisting of >> bugfixes). > > I would be big +1 to try to break out of the slow release train and > allow ourselves to cut 3.5, 3.6, .... quickly. Before I commit a > fix for MATH-1213 though, can we agree with what we are about to cut > is 3.5?
If we release minor releases more often I have no objection. Otherwise it feels wrong to release 3.5 just to get an urgent bugfix out of the door. Thomas --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
