Hi Georg. I don't want to get into a flaming debate about versioning. I've seen on the internet what happens.
> So to me it really boils down to not counting down when moving forward for people upgrading the health checks - it would feel awkward if you had 1.0.2 and then you use 1.0.0 again It makes no sense to me, sorry. The namespace is different, it's a different lib. The migration path would not be affected at all by this. I've seen multiple libraries migrate and reset the version space, which makes a lot of sense. (To me, actually having the same version as the sling one would be confusing if I needed to migrate, but that's my own view. The only libs I've seen that do that are libs with mocks) > Also the Sling API has been stable for long (Bertrand has done a great job initially designing it), why would it be all the sudden not good anymore? It's at 2.0.0, so well... 1.0.0 wasn't perfect :). But joke aside... I expect the need to change the API. What's the purpose of moving the HCs to Felix? Broadening the scope, right? This comes with different needs than before. As Ray said, it's weird. Someone might be tempted to release versions in sync of the one in sling and the one in felix and that would be a big no-no for me. - Andrei On Fri, Feb 1, 2019 at 2:07 PM Georg Henzler <fe...@ghenzler.de> wrote: > > I'm taking the hat of the many developers out there in the world that > use the API (the ones we cannot talk to and that do not read this > mailing list) - 2.0.0 it is to not confuse them. See my response from > this morning [1] > > Also the Sling API has been stable for long (Bertrand has done a great > job initially designing it), why would it be all the sudden not good > anymore? > > -Georg > > [1] https://www.mail-archive.com/dev@felix.apache.org/msg47690.html > > On 2019-02-01 14:00, Andrei Dulvac wrote: > > Hi, > > > > (Unfortunately I haven't had a chance to look over it properly) > > > > I'd just like to say I agree with Christian and Ray - a 2.0.0 initial > > release is not only weird, but confusing. Why? > > I support doing 0.X.0 releases until the api is proven to be stable and > > then release *1.0.0*. Why 2.0.0? > > > > - Andrei > > > > > > > > On Fri, Feb 1, 2019 at 1:53 PM Georg Henzler <fe...@ghenzler.de> wrote: > > > >> Hi Bertrand, > >> > >> > Testing that on snapshots is not optimal IMO as those are potentially > >> > moving targets. > >> > >> Let's do it like this then: I'll push a release 0.1.0 today (with > >> another short [VOTE] today that I would ask you to quickly vote +1 > >> for) > >> and I'll leave this vote and [1] open until Christian and you had the > >> chance to test more in detail with non-SNAPSHOT 0.1.0 artifacts. > >> > >> End of next week we can then close this vote if there is no good > >> reason > >> to cancel it. > >> > >> -Georg > >> > >> [1] > >> https://repository.apache.org/content/repositories/orgapachefelix-1279/ > >> > >> On 2019-02-01 12:59, Bertrand Delacretaz wrote: > >> > On Thu, Jan 31, 2019 at 5:37 PM Christian Schneider > >> > <ch...@die-schneider.net> wrote: > >> >> ...How about releasing 0.1.0 now and release a 2.0.0 in two weeks? > >> >> It would give people time to test the new project and still allow us > >> >> to do > >> >> incompatible changes.... > >> > > >> > I'm also strongly in favor of that, especially as these modules > >> > migrated from Sling and people will expect backwards compatibility. > >> > > >> > Testing that on snapshots is not optimal IMO as those are potentially > >> > moving targets. > >> > > >> > Releases are cheap - making another 1.0.0 or 2.0.0 release soon > >> > shouldn't be a problem. > >> > > >> > I'm -0 on releasing these modules as 2.0.0. > >> > > >> > -Bertrand > >> >
