Hi Jan,
Looking at the schedule[1] today I am confused by one thing, Fedora policy states that any slip pushes out all other milestones. which would mean Beta Freeze is next week not this week.
Dennis
On Mon, 2016-04-18 at 15:11 -0500, Dennis Gilmore wrote:
Hi Jan,
Looking at the schedule[1] today I am confused by one thing, Fedora policy states that any slip pushes out all other milestones. which would mean Beta Freeze is next week not this week.
Dennis
Er. Yes. That's a point.
I tend to just take it on faith that the schedules get adjusted after slips, but AFAICS, the F24 schedule was not adjusted after the Alpha slip. https://fedoraproject.org/w/index.php?title=Releases/24/Schedule&diff=43... is when the wiki schedule was adjusted for the Alpha slip, but the only date that was changed was the Alpha release date, no other dates were touched.
As things stand it's possible we could make the non-adjusted Beta dates, though there's some missing test coverage we'd really need to get to, but it's definitely not what we've done before without an explicit decision to *not* change the later dates.
On Mon, Apr 18, 2016 at 10:30 PM, Adam Williamson adamwill@fedoraproject.org wrote:
On Mon, 2016-04-18 at 15:11 -0500, Dennis Gilmore wrote:
Hi Jan,
Looking at the schedule[1] today I am confused by one thing, Fedora policy states that any slip pushes out all other milestones. which would mean Beta Freeze is next week not this week.
Dennis
Er. Yes. That's a point.
I tend to just take it on faith that the schedules get adjusted after slips, but AFAICS, the F24 schedule was not adjusted after the Alpha slip. https://fedoraproject.org/w/index.php?title=Releases/24/Schedule&diff=43... is when the wiki schedule was adjusted for the Alpha slip, but the only date that was changed was the Alpha release date, no other dates were touched.
As things stand it's possible we could make the non-adjusted Beta dates, though there's some missing test coverage we'd really need to get to, but it's definitely not what we've done before without an explicit decision to *not* change the later dates.
I have only changed the Alpha release date and intentionally have left the Beta and Final the same. This has been communicated to Marketing, unfortunately (my fault) not to rel-eng nor QA. The reason was mainly not to affect F25. Is it acceptable to have Beta & Final release dates for F24 as stated on [1], or we really need to slip these dates ?
[1] https://fedoraproject.org/wiki/Releases/24/Schedule
Regards, Jan
-- Adam Williamson Fedora QA Community Monkey IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net http://www.happyassassin.net
On Mon, 2016-04-18 at 22:41 +0200, Jan Kurik wrote:
On Mon, Apr 18, 2016 at 10:30 PM, Adam Williamson adamwill@fedoraproject.org wrote:
On Mon, 2016-04-18 at 15:11 -0500, Dennis Gilmore wrote:
Hi Jan,
Looking at the schedule[1] today I am confused by one thing, Fedora policy states that any slip pushes out all other milestones. which would mean Beta Freeze is next week not this week.
Dennis
Er. Yes. That's a point.
I tend to just take it on faith that the schedules get adjusted after slips, but AFAICS, the F24 schedule was not adjusted after the Alpha slip. https://fedoraproject.org/w/index.php?title=Releases/24/Schedule&di ff=439901&oldid=435873 is when the wiki schedule was adjusted for the Alpha slip, but the only date that was changed was the Alpha release date, no other dates were touched.
As things stand it's possible we could make the non-adjusted Beta dates, though there's some missing test coverage we'd really need to get to, but it's definitely not what we've done before without an explicit decision to *not* change the later dates.
I have only changed the Alpha release date and intentionally have left the Beta and Final the same. This has been communicated to Marketing, unfortunately (my fault) not to rel-eng nor QA. The reason was mainly not to affect F25. Is it acceptable to have Beta & Final release dates for F24 as stated on [1], or we really need to slip these dates ?
From QA's side we can at least try to make the unamended date, as I said.
On Mon, Apr 18, 2016 at 01:47:49PM -0700, Adam Williamson wrote:
Is it acceptable to have Beta & Final release dates for F24 as stated on [1], or we really need to slip these dates ?
From QA's side we can at least try to make the unamended date, as I said.
Awesome. I'd really, really not like to slip F25 further, even if F24 does end up with an adjustment. That's because that's targetted at Nov 8 (https://fedoraproject.org/wiki/Releases/25/Schedule) already, and that gives us one week of slip before we're in Thanksgiving week, and if that happens, we're probably not really releasing the week after either, leaving with another December release (and people worrying that the F26 schedule can't possibly target May).
On Monday, April 18, 2016 5:30:49 PM CDT Matthew Miller wrote:
On Mon, Apr 18, 2016 at 01:47:49PM -0700, Adam Williamson wrote:
Is it acceptable to have Beta & Final release dates for F24 as stated on [1], or we really need to slip these dates ?
From QA's side we can at least try to make the unamended date, as I said.
Awesome. I'd really, really not like to slip F25 further, even if F24 does end up with an adjustment. That's because that's targetted at Nov 8 (https://fedoraproject.org/wiki/Releases/25/Schedule) already, and that gives us one week of slip before we're in Thanksgiving week, and if that happens, we're probably not really releasing the week after either, leaving with another December release (and people worrying that the F26 schedule can't possibly target May).
I think we have all of the releng changes lined up for Beta, though another week would let us get layered image build service in production. I do not think that pushing 24 out an extra week needs to delay f25 in any way shape or form. Mostly I was worried we had not followed policy and we have perhaps not left enough time for developers to get their Beta fixes in.
Dennis
On Mon, Apr 18, 2016 at 06:36:02PM -0500, Dennis Gilmore wrote:
I think we have all of the releng changes lined up for Beta, though another week would let us get layered image build service in production. I do not think that pushing 24 out an extra week needs to delay f25 in any way shape or form. Mostly I was worried we had not followed policy and we have perhaps not left enough time for developers to get their Beta fixes in.
Also awesome, and I agree with your worry.
On Tue, Apr 19, 2016 at 1:36 AM, Dennis Gilmore dennis@ausil.us wrote:
On Monday, April 18, 2016 5:30:49 PM CDT Matthew Miller wrote:
On Mon, Apr 18, 2016 at 01:47:49PM -0700, Adam Williamson wrote:
Is it acceptable to have Beta & Final release dates for F24 as stated on [1], or we really need to slip these dates ?
From QA's side we can at least try to make the unamended date, as I said.
Awesome. I'd really, really not like to slip F25 further, even if F24 does end up with an adjustment. That's because that's targetted at Nov 8 (https://fedoraproject.org/wiki/Releases/25/Schedule) already, and that gives us one week of slip before we're in Thanksgiving week, and if that happens, we're probably not really releasing the week after either, leaving with another December release (and people worrying that the F26 schedule can't possibly target May).
I think we have all of the releng changes lined up for Beta, though another week would let us get layered image build service in production. I do not think that pushing 24 out an extra week needs to delay f25 in any way shape or form. Mostly I was worried we had not followed policy and we have perhaps not left enough time for developers to get their Beta fixes in.
Thanks Dennis for your readiness as well as for opening this topic. I will take this as a lesson learnt and will communicate such deviation from policy in a better way if there will be a need to do such things in the future.
Regards, Jan
Dennis
On Tuesday, April 19, 2016 1:25:50 PM CDT Jan Kurik wrote:
On Tue, Apr 19, 2016 at 1:36 AM, Dennis Gilmore dennis@ausil.us wrote:
On Monday, April 18, 2016 5:30:49 PM CDT Matthew Miller wrote:
On Mon, Apr 18, 2016 at 01:47:49PM -0700, Adam Williamson wrote:
Is it acceptable to have Beta & Final release dates for F24 as stated on [1], or we really need to slip these dates ?
From QA's side we can at least try to make the unamended date, as I said.
Awesome. I'd really, really not like to slip F25 further, even if F24 does end up with an adjustment. That's because that's targetted at Nov 8 (https://fedoraproject.org/wiki/Releases/25/Schedule) already, and that gives us one week of slip before we're in Thanksgiving week, and if that happens, we're probably not really releasing the week after either, leaving with another December release (and people worrying that the F26 schedule can't possibly target May).
I think we have all of the releng changes lined up for Beta, though another week would let us get layered image build service in production. I do not think that pushing 24 out an extra week needs to delay f25 in any way shape or form. Mostly I was worried we had not followed policy and we have perhaps not left enough time for developers to get their Beta fixes in.
Thanks Dennis for your readiness as well as for opening this topic. I will take this as a lesson learnt and will communicate such deviation from policy in a better way if there will be a need to do such things in the future.
I suggest going forward we follow policy unless FESCo decides to make an exception. People know how long to expect between releases. I am not sure why you decided to not follow policy, but that decision has to come from FESCo not from you. and needs to be communicated with the whole Fedora community. given the reminder on Fedora media Reminder we are not close to ready now. and really could use the extra week to try get things sorted out and lined up.
Dennis
-
Adam Williamson -
Dennis Gilmore -
Jan Kurik -
Matthew Miller