Re: [Tails-dev] Release versioning

Delete this message

Reply to this message
Autor: sajolida
Data:  
Para: The Tails public development discussion list
Assunto: Re: [Tails-dev] Release versioning
anonym:
> intrigeri:
>> intrigeri wrote (18 Aug 2015 11:09:35 GMT) :
>>> Please do whatever's needed for the change you proposed in our various
>>> repos + Redmine etc., after leaving some time to others to comment
>>> further. Perhaps we should make the final decision at the
>>> September meeting.
>>
>> I can't find a follow-up to this discussion.


There was none. I didn't hear back from "we should make the final
decision at the September meeting" and pinged anonym in private a couple
of times.

>> To sum up the options we have:
>>
>> A. Keep "odd for major release / even for minor ones" in general but
>>    call the first Tails/Jessie release 2.0. Call the next one 2.1
>>    because:
>>    - it'll likely introduce tons of changes to fix bugs in 2.0 anyway
>>    - 2.0 will be feature-frozen in 2 weeks, so IMO it makes sense to
>>      have a release with new features allowed in March

>
> Just wanted to say that I agree to those two points...


So you are proposing to have two major releases in a row: January 26 and
March 8. This goes beyond mere release versioning so let's agree on this
first and then adjust release numbers as we like.

Major releases are different minor releases as they imply an RC and more
work scheduled for the RM. It seems like both of you agree on doing a
major release on March 8. I personally don't mind.

>> B. Switch to "even for major releases / odd for minor ones", call the
>>    first Tails/Jessie release 2.0, and then
>>    1. find out how to deal with no release with new features allowed
>>       between November and April; exceptionally allow new features in
>>       2.1?

>
> ... but that's easily achievable by skipping 2.1, i.e. we plan 2.2 six
> weeks after 2.0 (in March). Both schemes suffer from the potential need
> of release skipping like this, as we already have noticed.


Yes, as long as we decide to have consistent semantic between "odd" and
"even" (I think intrigeri was strongly in favor of this) while allowing
ourselves to put out two majors releases in a row (you're proposing this
right now), then all number scheme that we can think of would require us
skipping numbers.

>>    2. adjust whatever is needed for this change

>
> I could do this, if we decide to switch.
>
>> As said already I can live with both, but I'd like to see a decision
>> made on this soon, and won't do the work required by B.
>
> I don't care much either. It does make sense that e.g. 2.0 should be a
> major release, and that is an even number, so that scheme has that
> additional consistency in its favour..


I agree with anonym. I don't mind doing the work needed for B. And I'm
fine with having 2.0 on January 26 and 2.2 on March 8.