On Mon, Apr 18, 2011 at 09:37:22PM +1000, Justin Clift wrote:
On 12/04/2011, at 3:36 AM, David Lutterkort wrote:
> On Mon, 2011-04-11 at 00:30 +1000, Justin Clift wrote:
>> For example, should we start with 0.1.0, or 0.5.0, 1.0.0,
>> or something else? (ie Aeolus 0.1.0)
>
> If you call it 0.1.0 Alpha, I'd expect there to be a 0.1.0 Beta, 0.1.0
> RC and 0.1.0 GA together with a dedicated cycle to get from one state to
> the other.
>
> If, as is common for open source projects, release qualification is
> simply best effort, rather than through a well-defined process, I
> wouldn't bother with alpha, beta etc. and just go straight to RC or even
> GA.
Good point. Alpha/beta/rc/etc naming is probably better suited
to a point in the future, where we have a number of stable releases
done.
Personally, for a consistent numbering approach, Jason's idea appeals to
me. :)
+ Each release from a major iteration increments a minor number
ie 0.1.0 -> 0.2.0
+ Each release from a sprint increments the revision number
ie 0.1.0 -> 0.1.1
If a royally bad bug turns up and _needs_ fixing immediately, we could
either just increase the revision number again (ie 0.1.1 -> 0.1.2)
or do a sub-revision thing (ie 0.1.1.1). Or maybe something else.
We can cross that bridge if/when we come to it. :)
Anyone have objections to using Jason's suggested approach? People ok
with it?
We could vote (+1/-1), but my suspicion is people don't really care
that strongly either way so far. ;>
We'll go with Jay's concept.
--Hugh
--
== Hugh Brock, hbrock(a)redhat.com ==
== Engineering Manager, Cloud BU ==
== Aeolus Project: Manage virtual infrastructure across clouds. ==
==
http://aeolusproject.org ==
"I know that you believe you understand what you think I said, but I’m
not sure you realize that what you heard is not what I meant."
--Robert McCloskey