[Emms-patches] Re: Version numbering scheme

[Lucas Bonnet]

Tim Landscheidt <address@hidden> writes:

> Lucas Bonnet <address@hidden> wrote:
>
>>>> Yeah, we should at least have two branches now:
>>>>  - 3.0-maintenance, to apply bugfixes to the released (and quite old)
>>>>    EMMS 3.0
>>>>  - HEAD, where new code is pushed
>>>>  - plus the temporary branches where fun stuff breaks everything, like
>>>>    typos-fix and so on, which should eventually get merged into HEAD.
>
>>>> I really need to get up to speed with git :)
>
>>> Oh, I didn't mean to talk about so low-level stuff as bran-
>>> ches; I'm thinking more about the user who doesn't "git
>>> clone" and still wants to know whether the snapshot he down-
>>> loaded from somewhere on the web is "3.0" or "3.0" :-).
>
>> Ah, right. We can start by changing the version numbers in HEAD to
>> something different. How about 3.5, which will be changed to 4.0 when we
>> do a release?
>
> I don't find picking arbitrary version numbers very help-
> ful :-). How would someone know that "3.5" is a "moving tar-
> get" and not a well-defined release, but "2.0", "2.1", "3.0"
> and "4.0" are?

By putting that into a doc somewhere in the release, presumably :) But I
get your point, it should probably be more obvious than that, and your
scheme is doing a good job of being obvious.

If no one else has something to say about that, we can work on this
scheme:
 - see the diffs between the old 3.0 and HEAD, and extract patches that
   are bugfixes, to make a 3.1 release
 - update the version number in HEAD to be 4.0dev
 - RELEASE!


Regards,
-- 
Lucas

pgpYjE3uL70lV.pgp
Description: PGP signature