Re: [cross-project-issues-dev] Question on Kepler SR1 release review

[Doug Schaefer <dschaefer@xxxxxxx>]

I guess we'll do that too for CDT but not announce the release until the SR and to avoid the confusion we won't mention the .1 in the version number. Again, our primary release vehicle is the EPP package. It's confusing enough well be doing a minor release in an SR package.

Sent from my BlackBerry 10 smartphone on the Rogers network.

From: Matthias Sohn

Sent: Wednesday, August 14, 2013 6:29 PM

To: Cross project issues

Reply To: Cross project issues

Subject: Re: [cross-project-issues-dev] Question on Kepler SR1 release review

On Wed, Aug 14, 2013 at 11:49 PM, David M Williams <david_williams@xxxxxxxxxx> wrote:

> was this decision mentioned on this list ? I wasn't aware of it and
> also can't find
> an announcement in the cross list's archive.
>
> This means we'll ship EGit 3.1 not with Kepler SR1 but from our own
> updates site
> since 1 month before SR1 already passed.

 

I meant 1 month before RC1

 

I'm not sure what to say. Sorry for not communicating better? Or, I assume Planning Council reps communicate with the projects they represent? Or, when did you think the deadline was?

I just want to learn how to prevent that I miss such decisions in the future.

And I'll try to find our PC rep to find out why he didn't tell us.

 

Or, should I say, remember we had to delay Juno SR2 by one week, for the first time ever, due to some last minute problems due to a project with late feature contributions, no rampdown plan, no incremental participation in release candidates, etc? That's why the rule was put in place. It is important we are on time, and it is important our quality improves each SR; no new problems introduced. That "rule" is the best concrete thing we could come up with to help ensure that. It was not a spur of the moment decision, but discussed over several meetings, with plenty of opportunity for Planning Council reps to solicit feedback and give advice.

If interested, some history in
https://bugs.eclipse.org/bugs/show_bug.cgi?id=401639
http://wiki.eclipse.org/Planning_Council/February_22_2013
http://wiki.eclipse.org/Planning_Council/March_06_2013#Juno_SR2
http://wiki.eclipse.org/Planning_Council/March_24_2013

thanks for the pointers

 

All that said, I should (again) remind everyone there is an "exception process"
http://wiki.eclipse.org/SimRel/Simultaneous_Release_Requirements#Planning_Council_Exception_Process
If you were missing the deadline by a day or two, I'm sure there'd be no problem getting an exception,
but if you are planning on releasing and contributing right before RC4 then pretty sure you would not get one.

There is no shame at all in releasing from your own project's repository on your own schedule, if that better suits your team. Its often a better approach for some.

that's exactly what we will do for the next EGit release 3.1. And I will create another pure maintenance release

to ship a few fixes with Kepler SR1.

But, to participate in Simultaneous Release requires something more ... we have to be well "coordinated" with each other, to meet the expectations of users and adopters.

So, I have decided ... I will just say, my personal apologies for not communicating this new rule  better, and in a more timely fashion. I understand how you might have made different plans if you had known about it earlier. I hope its not too much extra work for you to adjust your plans, one way or another. Thanks for your contributions to Eclipse.

yeah, we had different plans, but, well, no big damage occurred,

we can change our plans to comply with the PC's decision.

--

Matthias