On 22.03.2013 15:38, Daniel Veillard wrote:
On Fri, Mar 22, 2013 at 02:23:34PM +0000, Daniel P. Berrange wrote:
> On Fri, Mar 22, 2013 at 10:17:07PM +0800, Daniel Veillard wrote:
>> On Fri, Mar 22, 2013 at 02:07:05PM +0000, Daniel P. Berrange wrote:
>>> On Fri, Mar 22, 2013 at 09:36:03PM +0800, Daniel Veillard wrote:
>>>> The 1.0.3 release was on the 5th March, and right now we have
>>>> accumulated 'only' 150 commits since the release. Based on this
I
>>>> would suggest to wait a couple of weeks before to enter a freeze
>>>> for the following release. This mean we drift from the usual end
>>>> of month, but the ratio of freeze/devel will remain more or less
>>>> constant as well as the expected size of change in the new release.
>>>> So if this is fine I would suggest to enter freeze for 1.0.4
>>>> on the 5th of April for a release around the 12. Unless there is
>>>> a reason to push a release earlier,
>>>
>>> I would prefer it if we just stuck to a release on/near April 1st
>>> regardless of how many changes have accumulated. IMHO a predictable
>>> release date once a month on/near 1st of the month is more important
>>> than the amount of code that has been changed.
>>
>> to release on the 1st we would have to freeze this Monday meaning
>> only 2.5 weeks of development after a freeze for 1.3 which took 1.5
>> weeks. I was considering the ratio of freeze time vs. open time too.
>
> I don't really think that's a problem myself, it is just to be
> expected due to the longer than usual 1.3 freeze. People have
> had just as much time to /write/ their code, all that changed
> was the time available to /merge/ their code into upstream.
>
> I'd like to see is be much stricter about following a monthly
> release schedule in general. It makes it easier for distro
> maintainers to accurately plan ahead for when libvirt releases
> will sync up with their schedules, if they can reliably know
> that we'll always release on the nearest weekday that follows
> the 1st of the month.
So that mean freezing on monday, I can do that !
Daniel
That effectively cuts a day for us, leaving us with only four days of
freeze since on 1st April it's Easter Monday => public holiday in most
countries (at least I've taken look at UK and Czech Republic). But I can
live with that since the current status of the HEAD looks good.
Michal