Refining the update queues/process [Was: Worthless updates]

[Kevin Kofler]

Seth Vidal wrote:
> If you can't understand it, perhaps you should reconsider your role on a
> technical committee like fesco.

I understood your sentence, that doesn't make it any less jargon.
 
>> * And most importantly, even if we were to accept that it could lead to
>> better QA (which I doubt), would it really be worth making users wait a
>> FULL MONTH for their updates and forcing them to pull a HUGE update in
>> one single transaction rather than spread over time? IMHO, HELL NO!
> 
> It will mean regularity and predictability. It will let users and admins
> alike know when they can expect new things. Doing it once a month means
> that they only need to cope with 13 of these per fedora release.

13 huge updates are a lot more pain for people with slow connections than 
many small ones. It'd also make Fedora effectively useless for those people 
like me who use it because of the frequent updates. It makes this look like 
M$ Patch Tuesdays instead of a fast-moving GNU/Linux distribution.

The solution to get more effective testing is to speed up pushes to testing 
(and the technical issues with that are already being discussed), not to 
slow down pushes to stable. Time is critical. Waiting a full month is only 
going to cause problems and not solve anything at all.

> Predictability means you can plan for effectively.

No. It means a user can't plan when to update anymore, you're forcing your 
schedule on the user! With the current system, the user can update on 
his/her OWN schedule, daily, weekly, monthly or whatever, and at whatever 
moment he/she wishes. With your M$-style system, they'd have to deal with 
monthly big dumps and no other choice.

>> Why? Because you say so? We aren't doing that stuff now and things are
>> working just fine, thank you very much! We don't HAVE to change anything
>> at all!
> 
> We will be doing it or something very much like it. Good luck with your
> fight.

I don't see how you can be this definite. There was far from a consensus 
that a time-based solution is the right thing to do even among those who 
opposed direct stable pushes, I was clearly not the only one opposed to 
that!

It's quite sad that you fail to realize how badly your proposal is flawed 
and what disastrous consequences its implementation would have. We'd be 
breaking Fedora for a huge portion of its current userbase!

        Kevin Kofler