Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-12-01 Thread Chris Holmes
Oh, one additional thought: It might make sense to differentiate between major and minor release managers. And indeed it may make sense to have the major release manager take responsibility from beta -> .0 release. I think we're mostly here talking about stable releases, but we should be clear a

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-12-01 Thread Chris Holmes
Hey all, apologies for sounding in late on this one, but it's been really nice to see all the discussion among the PSC. In general I've been the one to most push for releasing often in the past, and I do think doing releases often is a Very Good Thing. I personally think the ideal is monthly rele

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-29 Thread Rob Atkinson
I think it would be a great idea to factor in cost for a release. It would be even better if there was a review+release capability - so party A could factor in party B to do a review and release (or even Party C to do the release). Maybe we should but a fixed fee $ (or euro :-) ) figure on a rel

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-27 Thread Andrea Aime
Jody Garnett ha scritto: >> I am still try to convince myself that the "release early, release >> often" is a good approach. My preference goes to "release when it's >> worth, release something that is rock solid" > > One solid point the in the favour of release early release often is it > takes s

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-26 Thread Jody Garnett
> I am still try to convince myself that the "release early, release > often" is a good approach. My preference goes to "release when it's > worth, release something that is rock solid" One solid point the in the favour of release early release often is it takes some of the pressure off the planni

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-26 Thread Jody Garnett
As a nod to simone; we could preform cite tests and the QA process on a nightly release and "mark it" as a *good* release for others. The eclipse project does this by tagging the occasional "stream integration build" that has passed integration testing; as opposed to their "stream nightly build".

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-26 Thread Jody Garnett
This discussion is going in interesting directions. For me the main benefit of releasing at a set interval (each month, each quarter, or every 6 months) is predictability. It allows users to plan their life with some measure of confidence (the nightly build fix worked; so the fix will be availabl

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-24 Thread Simone Giannecchini
On Tue, Nov 24, 2009 at 12:58 PM, Andrea Aime wrote: > Simone Giannecchini ha scritto: >> My perspective is different. To me saying that we should make a >> release once every month is different than saying we should gather >> feedback each month on whether to make or not a release. >> If I were a

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-24 Thread Andrea Aime
Simone Giannecchini ha scritto: > My perspective is different. To me saying that we should make a > release once every month is different than saying we should gather > feedback each month on whether to make or not a release. > If I were a user, if we went for approach a> I would expect a release >

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-24 Thread Simone Giannecchini
Ciao Andrea, please, read below... --- Ing. Simone Giannecchini GeoSolutions S.A.S. Founder - Software Engineer Via Carignoni 51 55041 Camaiore (LU) Italy phone: +39 0584983027 fax: +39 0584983027 mob:+39 333 8128928 http://www.geo-s

Re: [Geoserver-devel] PSC: GSIP 43, release cycle lenght

2009-11-24 Thread Andrea Aime
Simone Giannecchini ha scritto: > Dear all, > providing some feedback. > > I think that a 1 month release cycle may be too much of a burden, > especially since we have nightly builds that people can use, therefore > I would suggest to reconsider this timing. ... > The key for me is having PSC de