With the 4.5 kernel out and the merge window for 4.6 opened up, we had to make a decision on what the release kernel for F24 would be. The decision has been made to ship F24 with the 4.5 kernel with 4.6 available as an update once it is ready. Timing wise, 4.6 *should* release just before the final freeze for F24, but that is cutting it insanely close. Should Fedora move on as scheduled, and 4.6 have some delay due to a bug that impacts users, that would be unfortunate. This means we have a good bit of time to make sure that everything is working as intended with 4.5 in Fedora. It also means that any installer critical fixes will need to be backported to 4.5.
Thanks, Justin
On 03/16/2016 04:01 PM, Justin Forbes wrote:
With the 4.5 kernel out and the merge window for 4.6 opened up, we had to make a decision on what the release kernel for F24 would be. The decision has been made to ship F24 with the 4.5 kernel with 4.6 available as an update once it is ready. Timing wise, 4.6 *should* release just before the final freeze for F24, but that is cutting it insanely close. Should Fedora move on as scheduled, and 4.6 have some delay due to a bug that impacts users, that would be unfortunate. This means we have a good bit of time to make sure that everything is working as intended with 4.5 in Fedora. It also means that any installer critical fixes will need to be backported to 4.5.
Given that 4.6 is out and current F24 final freeze is not scheduled until 2016-05-31 should not F24 be released with the 4.6 kernel?
JBG
On Tue, May 17, 2016 at 8:57 AM, Jóhann B. Guðmundsson johannbg@gmail.com wrote:
On 03/16/2016 04:01 PM, Justin Forbes wrote:
With the 4.5 kernel out and the merge window for 4.6 opened up, we had to make a decision on what the release kernel for F24 would be. The decision has been made to ship F24 with the 4.5 kernel with 4.6 available as an update once it is ready. Timing wise, 4.6 *should* release just before the final freeze for F24, but that is cutting it insanely close. Should Fedora move on as scheduled, and 4.6 have some delay due to a bug that impacts users, that would be unfortunate. This means we have a good bit of time to make sure that everything is working as intended with 4.5 in Fedora. It also means that any installer critical fixes will need to be backported to 4.5.
Given that 4.6 is out and current F24 final freeze is not scheduled until 2016-05-31 should not F24 be released with the 4.6 kernel?
I think the original logic is still sound. There have been three delays for Fedora 24 already, in the original schedule today was to be GA. I don't think it's worth any risk for another slip.
On 05/17/2016 04:32 PM, Chris Murphy wrote:
On Tue, May 17, 2016 at 8:57 AM, Jóhann B. Guðmundsson johannbg@gmail.com wrote:
On 03/16/2016 04:01 PM, Justin Forbes wrote:
With the 4.5 kernel out and the merge window for 4.6 opened up, we had to make a decision on what the release kernel for F24 would be. The decision has been made to ship F24 with the 4.5 kernel with 4.6 available as an update once it is ready. Timing wise, 4.6 *should* release just before the final freeze for F24, but that is cutting it insanely close. Should Fedora move on as scheduled, and 4.6 have some delay due to a bug that impacts users, that would be unfortunate. This means we have a good bit of time to make sure that everything is working as intended with 4.5 in Fedora. It also means that any installer critical fixes will need to be backported to 4.5.
Given that 4.6 is out and current F24 final freeze is not scheduled until 2016-05-31 should not F24 be released with the 4.6 kernel?
I think the original logic is still sound. There have been three delays for Fedora 24 already, in the original schedule today was to be GA. I don't think it's worth any risk for another slip.
So you prefer not catching potentially any bugs before final GA release but rather expose them to the end users through 0 day update instead?
JBG
On Tue, May 17, 2016 at 11:05 AM, Jóhann B. Guðmundsson johannbg@gmail.com wrote:
On 05/17/2016 04:32 PM, Chris Murphy wrote:
On Tue, May 17, 2016 at 8:57 AM, Jóhann B. Guðmundsson johannbg@gmail.com wrote:
On 03/16/2016 04:01 PM, Justin Forbes wrote:
With the 4.5 kernel out and the merge window for 4.6 opened up, we had to make a decision on what the release kernel for F24 would be. The decision has been made to ship F24 with the 4.5 kernel with 4.6 available as an update once it is ready. Timing wise, 4.6 *should* release just before the final freeze for F24, but that is cutting it insanely close. Should Fedora move on as scheduled, and 4.6 have some delay due to a bug that impacts users, that would be unfortunate. This means we have a good bit of time to make sure that everything is working as intended with 4.5 in Fedora. It also means that any installer critical fixes will need to be backported to 4.5.
Given that 4.6 is out and current F24 final freeze is not scheduled until 2016-05-31 should not F24 be released with the 4.6 kernel?
I think the original logic is still sound. There have been three delays for Fedora 24 already, in the original schedule today was to be GA. I don't think it's worth any risk for another slip.
So you prefer not catching potentially any bugs before final GA release but rather expose them to the end users through 0 day update instead?
I refuse the premise that the kernel team is going to release a 4.6.x kernel that isn't ready as a 0 day update.
On 05/17/2016 05:25 PM, Chris Murphy wrote:
I refuse the premise that the kernel team is going to release a 4.6.x kernel that isn't ready as a 0 day update.
There is more extensive testing performed before GA release then there is in the update process hence what get's shipped in the GA release has better testing coverage regardless of people believes in Red Hat's kernel team hence it's better to ship the 4.6 in the final then to deliver it as an 0 day update.
JBG
On Tue, May 17, 2016 at 17:44:43 +0000, "Jóhann B. Guðmundsson" johannbg@gmail.com wrote:
On 05/17/2016 05:25 PM, Chris Murphy wrote:
I refuse the premise that the kernel team is going to release a 4.6.x kernel that isn't ready as a 0 day update.
There is more extensive testing performed before GA release then there is in the update process hence what get's shipped in the GA release has better testing coverage regardless of people believes in Red Hat's kernel team hence it's better to ship the 4.6 in the final then to deliver it as an 0 day update.
If the plan is to release with 4.6 as a zero day update, then it would seem that getting in now, before final freeze would be better. If it is going to just be an eventual update, but not necessarily available on release day, then not rushing things seems better.
On Tue, May 17, 2016 at 11:44 AM, Jóhann B. Guðmundsson johannbg@gmail.com wrote:
On 05/17/2016 05:25 PM, Chris Murphy wrote:
I refuse the premise that the kernel team is going to release a 4.6.x kernel that isn't ready as a 0 day update.
There is more extensive testing performed before GA release then there is in the update process hence what get's shipped in the GA release has better testing coverage regardless of people believes in Red Hat's kernel team hence it's better to ship the 4.6 in the final then to deliver it as an 0 day update.
You're the only person suggesting it would be a zero day update, rather than following the usual kernel rebasing process. The 4.6 kernel is mainline, not stable. I can't recall Fedora shipping with a mainline kernel. If testing coverage is your concern, 4.5.3 will have had more coverage by a lot than 4.6.0 will get in the next two weeks before freeze.
And then what? You expect the kernel team to carry backports for 4.6.0 for a week or two after GA? I don't see the point at all.
Historically it's about a month before a kernel goes stable, and then we'd see 4.6.1 right about when Fedora 24 goes GA, and 4.6.1 would still spend some time in u-t before it'd get to most users. I don't know why you think it'd be a zero day update.
On Tue, May 17, 2016 at 1:22 PM, Chris Murphy lists@colorremedies.com wrote:
I can't recall Fedora shipping with a mainline kernel.
Ha, well that's obviously wrong. Often Fedora has rc's in alpha and ships with that released kernel. So never mind on that point!
In any case it's up to the kernel team. I don't think anyone else's two cents really matters.
On Tue, May 17, 2016 at 05:44:43PM +0000, Jóhann B. Guðmundsson wrote:
There is more extensive testing performed before GA release then there is in the update process hence what get's shipped in the GA release has better testing coverage regardless of people believes in Red Hat's kernel team hence it's better to ship the 4.6 in the final then to deliver it as an 0 day update.
I have faith in the kernel and QA teams' regular testing and updates process. But there's another key difference you're leaving out: if a kernl problem arises at the last minute which affects the install media (anaconda or otherwise), that may involve significant work and delay. If a problem is discovered with a kernel in updates-testing, it's just the kernel.
I think Josh wrote nice blog about "kernel release cycle":
http://jwboyer.livejournal.com/51935.html
Based on it, I'd be surprised to see 4.6 land in F24 prior stable release.
Vít
Dne 18.5.2016 v 17:27 Matthew Miller napsal(a):
On Tue, May 17, 2016 at 05:44:43PM +0000, Jóhann B. Guðmundsson wrote:
There is more extensive testing performed before GA release then there is in the update process hence what get's shipped in the GA release has better testing coverage regardless of people believes in Red Hat's kernel team hence it's better to ship the 4.6 in the final then to deliver it as an 0 day update.
I have faith in the kernel and QA teams' regular testing and updates process. But there's another key difference you're leaving out: if a kernl problem arises at the last minute which affects the install media (anaconda or otherwise), that may involve significant work and delay. If a problem is discovered with a kernel in updates-testing, it's just the kernel.
-
Bruno Wolff III -
Chris Murphy -
Justin Forbes -
Jóhann B. Guðmundsson -
Matthew Miller -
Vít Ondruch