Re: [GRASS-PSC] [GRASS-dev] RFC 4: Release procedure

[Moritz Lennert]

On 20/10/15 09:18, Martin Landa wrote:

2015-10-20 9:12 GMT+02:00 Martin Landa :

Great, so I'll make a motion on this in the next days unless any
objection pops up before.


it seems that there are no objections :-) Martin


reading it again, I would say that in some cases upcoming RC2 could be
released as final. Eg. for upcoming 7.0.2 we have RC1 released. In two
weeks should become RC2, in this case I can image to release final
instead of RC2. So alternative roadmap could be:

Step 3 (X+30 days) - Hard freeze & RC1
Step 4 - Bug squashing
Step 5 (X+44 days) - Final.

In other words I would make RC2 as optional step only we will need it.
It can happen that two RCs could be overkill. What do you think?


The idea of the RC2 was to provoke some more last-minute testing as some 
fixes might have been introduced after RC1 and I'm not sure how many 
people test the release branch between RC's. This way we make it more 
prominent and can send out a call to everyone to please test RC2. This 
does not mean that RC2 cannot be identical to final. It's just a last 
chance to spot any serious issues.


So, I would plead for leaving it in. 5 days more or less is not that 
much, or ?


Moritz
___
grass-psc mailing list
grass-psc@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-psc

Re: [GRASS-PSC] [GRASS-dev] RFC 4: Release procedure

[Martin Landa]

Hi Markus,

2015-03-04 20:29 GMT+01:00 Markus Neteler :

[...]

> Great, so I'll make a motion on this in the next days unless any
> objection pops up before.

it seems that there are no objections :-) Martin

-- 
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa
___
grass-psc mailing list
grass-psc@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-psc

Re: [GRASS-PSC] [GRASS-dev] RFC 4: Release procedure

[Anna Petrášová]

On Tue, Oct 20, 2015 at 4:29 PM, Martin Landa 
wrote:

> Hi,
>
> 2015-10-20 9:36 GMT+02:00 Moritz Lennert :
>
> > The idea of the RC2 was to provoke some more last-minute testing as some
> > fixes might have been introduced after RC1 and I'm not sure how many
> people
> > test the release branch between RC's. This way we make it more prominent
> and
> > can send out a call to everyone to please test RC2. This does not mean
> that
> > RC2 cannot be identical to final. It's just a last chance to spot any
> > serious issues.
> >
> > So, I would plead for leaving it in. 5 days more or less is not that
> much,
> > or ?
>
> I understand the point, on the other hand it's extra work for release
> manager and packager which would sometimes make sense to avoid and be
> so not strict in the way that RC2 step could be optional (or skipped
> if no objection from community). On the other hand we can add more
> steps (RC3, RC4, ...) if it will be necessary in the case of extra
> complicated release.
>
> Any comments, ideas? Thanks, Martin
>

I agree with Martin, I guess it's quite a bit of work involved in it and it
seems we now started to release more often than in previous years, which is
a good trend. So I would rather release more often with less RCs.

Anna

>
> --
> Martin Landa
> http://geo.fsv.cvut.cz/gwiki/Landa
> http://gismentors.cz/mentors/landa
> ___
> grass-dev mailing list
> grass-...@lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-dev
>
___
grass-psc mailing list
grass-psc@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-psc

Re: [GRASS-PSC] [GRASS-dev] RFC 4: Release procedure

[Martin Landa]

Hi,

2015-10-20 9:36 GMT+02:00 Moritz Lennert :

> The idea of the RC2 was to provoke some more last-minute testing as some
> fixes might have been introduced after RC1 and I'm not sure how many people
> test the release branch between RC's. This way we make it more prominent and
> can send out a call to everyone to please test RC2. This does not mean that
> RC2 cannot be identical to final. It's just a last chance to spot any
> serious issues.
>
> So, I would plead for leaving it in. 5 days more or less is not that much,
> or ?

I understand the point, on the other hand it's extra work for release
manager and packager which would sometimes make sense to avoid and be
so not strict in the way that RC2 step could be optional (or skipped
if no objection from community). On the other hand we can add more
steps (RC3, RC4, ...) if it will be necessary in the case of extra
complicated release.

Any comments, ideas? Thanks, Martin

-- 
Martin Landa
http://geo.fsv.cvut.cz/gwiki/Landa
http://gismentors.cz/mentors/landa
___
grass-psc mailing list
grass-psc@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-psc