[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: the short road from rc1 to rc2 to lenny



Hi, 

some more context from irc:

<h01ger> probably better than reverting everything in trunk :-P
<h01ger> (well not better for me, its way more work, but so be it)
<h01ger> (the squeeze freeze is still planned for march atm)
<h01ger> pere, so far everybody i spoke too agreed on not pushing new stuff 
anymore and instead just keeping it like it is and release what we have, 
modulo p1+2 bugs and translations. constantly pushing new features prevents 
us from releasing. i find this *DEEPLY* frustrating, so i will go with this 
branch and not let anything into lenny matching these criteria. another 
pointrelease after our lenny release is fine, but we *need* to get to this 
release first. 
<pere> h01ger: I believe our view of the distribution quality differ slightly, 
and have tried to explain my view in a reply to your email on the list.
<h01ger> pere, it seems you're alone. the rest wants to release asap
<pere> btw, perhaps we should plan a date for the next point release.  I 
suspect we will discover lots of bugs when we release, and suggest we plan a 
point release 1-2 months after r0.
<h01ger> pere, that is the plan. but its kinda pointless if we dont stick to 
the previous step first, which is to release in the first place
<pere> for me, it only make sense to delay features that can be sensibly 
activated during upgrades.  the wpad and nagios change can be handled like 
that.  the default home page setting can't.
<h01ger> pere, so for other features you want to delay the release even more? 
<h01ger> that doesnt compute
<pere> h01ger: are you aware of other features we need that can't be 
implemented sensible during upgrades?
<pere> h01ger: and there is really no need to delay the release that I am 
aware of.
<h01ger> anyway, i'll go with the lenny branch approach now and will only care 
about trunk again, once lenny is out
<pere> h01ger: that seem like a bad approach to me.
<h01ger> pere, definitly no need, but lots of risks. also new features 
distract from PR, writing docs and testing. plus they introduce regressions
<h01ger> pere, i was asked by the project to do the release and define the 
rules for it. here we go. enjoy your ride

If the project wants me to stop working on the release, please say so, I'd 
have no problem on working on the release of Debian squeeze instead.


cheers,
	Holger

Attachment: signature.asc
Description: This is a digitally signed message part.


Reply to: