On 5 May 2009, at 13:20, Martin Aspeli wrote:
Yes, I agree with all of these, and do think they're needed.
So, 3.5 is a compromise. The skipping of 3.4 actually helps back the
story up. We could try something else, like Plone 2009, but I'm
pretty sure we'd regret that in 2010 for one reason or another. And
PyPI wouldn't like it.
That's true, I am very nervous about the Plone 3.x compatibility
becoming >=3.0,<3.5-dev though - it is a change in policy to what
we've previously said and I do think it would catch people out.
I had some issues with mails bouncing as I'm subscribed with different
addresses to different lists, so, to reiterate, if we keep Plone 3.3
supported and make it clear that there's a significant difference
between 3.3 and 3.5, then I'm +0. I'd only be +1 if we could make
that explicit in the name. This is not because of the proposal, I'm
very much in favour of the proposal in principle, it's just the fact
we lose our easy-to-understand compatibility promise by making a
decision very much like the one we made in Plone 2.5.
Yeah, thanks for helping. :p
There may well not be a name that would satisfy me. I can't think of
one, you can't think of one, there may just not be one.
I think that'd be the case under the "two supported versions" policy.
But that would mean that when 4.0 comes out 3.<5 becomes unsupported
completely. Is that desirable?
I'm not trying to be an arse here Martin, really, I'm just concerned
about 3.x users. We've managed to remove a lot of fear from upgrades
in the 3.x series, I don't want to lose that.
Framework-Team mailing list