Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-30 Thread Joe Conway 
On 03/30/2017 10:59 AM, Magnus Hagander wrote:
> On Thu, Mar 30, 2017 at 4:40 PM, Tom Lane  > wrote:
> 
> Robert Haas mailto:robertmh...@gmail.com>>
> writes:
> > On Thu, Mar 30, 2017 at 9:41 AM, Bruce Momjian  > wrote:
> >> I propose we go for a week delay in closing the commit fest, and we
> >> decide right now.  Ideally I like to to see delay in one-week 
> increments
> >> _and_ announce that a week before each deadline.
> 
> > Summary of opinions on this thread:
> 
> > - Tom thinks the RMT should consider extending by a day or two.
> 
> FWIW, while I'm sure we had a reason (back at the Ottawa 2016 meeting)
> for limiting the final 'fest to just a month, I'm quite sure we did
> not allow for most of the senior hackers being at a conference during
> the last week of that time.  So I now concur with Bruce's suggestion
> that a week's delay would be suitable.
> 
> 
> +1.

+1


-- 
Crunchy Data - http://crunchydata.com
PostgreSQL Support for Secure Enterprises
Consulting, Training, & Open Source Development



signature.asc
Description: OpenPGP digital signature

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-30 Thread Alvaro Herrera 
Robert Haas wrote:

> - Alvaro proposes allowing more time next time, but not to change the
> dates for this time.

FWIW I didn't realize that the NY conference was ongoing, so count me
for postponing the end of the current CF.

-- 
Álvaro Herrerahttps://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-30 Thread Magnus Hagander 
On Thu, Mar 30, 2017 at 4:40 PM, Tom Lane  wrote:

> Robert Haas  writes:
> > On Thu, Mar 30, 2017 at 9:41 AM, Bruce Momjian  wrote:
> >> I propose we go for a week delay in closing the commit fest, and we
> >> decide right now.  Ideally I like to to see delay in one-week increments
> >> _and_ announce that a week before each deadline.
>
> > Summary of opinions on this thread:
>
> > - Tom thinks the RMT should consider extending by a day or two.
>
> FWIW, while I'm sure we had a reason (back at the Ottawa 2016 meeting)
> for limiting the final 'fest to just a month, I'm quite sure we did
> not allow for most of the senior hackers being at a conference during
> the last week of that time.  So I now concur with Bruce's suggestion
> that a week's delay would be suitable.
>

+1.

-- 
 Magnus Hagander
 Me: http://www.hagander.net/
 Work: http://www.redpill-linpro.com/

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-30 Thread Petr Jelinek 
On 30/03/17 16:54, Alvaro Herrera wrote:
> Robert Haas wrote:
> 
>> - Alvaro proposes allowing more time next time, but not to change the
>> dates for this time.
> 
> FWIW I didn't realize that the NY conference was ongoing, so count me
> for postponing the end of the current CF.
> 

+1, the conference makes it bit inconvenient to have freeze on exactly 31st.

-- 
  Petr Jelinek  http://www.2ndQuadrant.com/
  PostgreSQL Development, 24x7 Support, Training & Services


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-30 Thread Stephen Frost 
* Tom Lane (t...@sss.pgh.pa.us) wrote:
> Robert Haas  writes:
> > On Thu, Mar 30, 2017 at 9:41 AM, Bruce Momjian  wrote:
> >> I propose we go for a week delay in closing the commit fest, and we
> >> decide right now.  Ideally I like to to see delay in one-week increments
> >> _and_ announce that a week before each deadline.
> 
> > Summary of opinions on this thread:
> 
> > - Tom thinks the RMT should consider extending by a day or two.
> 
> FWIW, while I'm sure we had a reason (back at the Ottawa 2016 meeting)
> for limiting the final 'fest to just a month, I'm quite sure we did
> not allow for most of the senior hackers being at a conference during
> the last week of that time.  So I now concur with Bruce's suggestion
> that a week's delay would be suitable.

Agreed, same here.

Thanks!

Stephen


signature.asc
Description: Digital signature

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-30 Thread Tom Lane 
Robert Haas  writes:
> On Thu, Mar 30, 2017 at 9:41 AM, Bruce Momjian  wrote:
>> I propose we go for a week delay in closing the commit fest, and we
>> decide right now.  Ideally I like to to see delay in one-week increments
>> _and_ announce that a week before each deadline.

> Summary of opinions on this thread:

> - Tom thinks the RMT should consider extending by a day or two.

FWIW, while I'm sure we had a reason (back at the Ottawa 2016 meeting)
for limiting the final 'fest to just a month, I'm quite sure we did
not allow for most of the senior hackers being at a conference during
the last week of that time.  So I now concur with Bruce's suggestion
that a week's delay would be suitable.

In past years this sort of decision would have been taken by -core,
and given the number of core members who have already weighed in
on this thread, I think Bruce's proposal would have passed easily.
But since we just delegated the setting of feature freeze to the RMT,
they are the ones who ought to do it.

regards, tom lane


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-30 Thread Robert Haas 
On Thu, Mar 30, 2017 at 9:41 AM, Bruce Momjian  wrote:
> I agree we need to extend, and not wait until any longer to do it.  We
> have people at the NYC conference and I don't want their memory of the
> conference being that they were stressed trying to work on closing the
> commit fest --- that could lead to bad memories and them declining
> future conference attendance.  If that delays our final release for a
> week, it is worth it.
>
> I propose we go for a week delay in closing the commit fest, and we
> decide right now.  Ideally I like to to see delay in one-week increments
> _and_ announce that a week before each deadline.

Summary of opinions on this thread:

- Tom thinks the RMT should consider extending by a day or two.
- Stephen agrees.
- Alvaro proposes allowing more time next time, but not to change the
dates for this time.
- Andres and Dave agree with Alvaro's surprise about the selected
date, but it's not clear whether they want to change things now or
just for next time.
- Robert thinks that the RMT shouldn't presume to vary the dates
without a broad consensus, and somewhat favors not extending.

Other views?

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-30 Thread Bruce Momjian 
On Tue, Mar 28, 2017 at 11:35:17PM -0400, Tom Lane wrote:
> Well, formally speaking, the RMT declares the onset of feature freeze.
> I expressed an opinion that they'd probably do so at the end of March,
> but it's their call as to exactly when to do so, and whether to grant
> any limited extensions/exceptions.
> 
> My own thought is that there's room for at least a few days' slop in
> the end date of the final commitfest, depending on what patches remain
> open and what the prospects are for getting them done.  (In the past
> we've sometimes let the final fest stretch on indefinitely, which is
> clearly the Wrong Thing; but that doesn't mean that the Right Thing is
> to say that it ends at 2017-04-01 00:00 UTC no matter what.)  The RMT
> should look at things in another day or two and make a judgment call
> about that.

I agree we need to extend, and not wait until any longer to do it.  We
have people at the NYC conference and I don't want their memory of the
conference being that they were stressed trying to work on closing the
commit fest --- that could lead to bad memories and them declining
future conference attendance.  If that delays our final release for a
week, it is worth it.

I propose we go for a week delay in closing the commit fest, and we
decide right now.  Ideally I like to to see delay in one-week increments
_and_ announce that a week before each deadline.

-- 
  Bruce Momjian  http://momjian.us
  EnterpriseDB http://enterprisedb.com

+ As you are, so once was I.  As I am, so you will be. +
+  Ancient Roman grave inscription +


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-29 Thread Alvaro Herrera 
Robert Haas wrote:
> On Wed, Mar 29, 2017 at 3:04 PM, Alvaro Herrera
>  wrote:

> > I was rather surprised to see the March commitfest declared to exactly
> > one month and feature freeze immediately thereafter.
> 
> That's true, but at
> https://wiki.postgresql.org/wiki/PgCon_2016_Developer_Meeting we
> agreed on March 31st as the feature freeze date.  I don't think the
> RMT should substantially vary a date that was agreed by a developer
> meeting with wide attendance; that seems like the tyranny of the few
> over the many.

Yes, I agree with that -- which is why I didn't comment when I read the
minutes several weeks ago.  I'm not proposing to change them, only
raising the point hoping that perhaps we'll choose differently next time.

-- 
Álvaro Herrerahttps://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-29 Thread Dave Page 
On Wed, Mar 29, 2017 at 3:10 PM, Andres Freund  wrote:
> On 2017-03-29 16:04:50 -0300, Alvaro Herrera wrote:
>> Tom Lane wrote:
>>
>> > My own thought is that there's room for at least a few days' slop in
>> > the end date of the final commitfest, depending on what patches remain
>> > open and what the prospects are for getting them done.  (In the past
>> > we've sometimes let the final fest stretch on indefinitely, which is
>> > clearly the Wrong Thing; but that doesn't mean that the Right Thing is
>> > to say that it ends at 2017-04-01 00:00 UTC no matter what.)  The RMT
>> > should look at things in another day or two and make a judgment call
>> > about that.
>>
>> I was rather surprised to see the March commitfest declared to exactly
>> one month and feature freeze immediately thereafter.  Last time around
>> we left 2 weeks between CF end and feature freeze; the previous one I
>> think we had the final CF last two months.  Not stretch on indefinitely,
>> but we know the final CF for a cycle takes more effort than previous
>> ones, so it seems reasonable to give more time.  We have a large number
>> of patches still waiting for review.
>
> +1

+1


-- 
Dave Page
Blog: http://pgsnake.blogspot.com
Twitter: @pgsnake

EnterpriseDB UK: http://www.enterprisedb.com
The Enterprise PostgreSQL Company


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-29 Thread Andres Freund 
On 2017-03-29 16:04:50 -0300, Alvaro Herrera wrote:
> Tom Lane wrote:
>
> > My own thought is that there's room for at least a few days' slop in
> > the end date of the final commitfest, depending on what patches remain
> > open and what the prospects are for getting them done.  (In the past
> > we've sometimes let the final fest stretch on indefinitely, which is
> > clearly the Wrong Thing; but that doesn't mean that the Right Thing is
> > to say that it ends at 2017-04-01 00:00 UTC no matter what.)  The RMT
> > should look at things in another day or two and make a judgment call
> > about that.
>
> I was rather surprised to see the March commitfest declared to exactly
> one month and feature freeze immediately thereafter.  Last time around
> we left 2 weeks between CF end and feature freeze; the previous one I
> think we had the final CF last two months.  Not stretch on indefinitely,
> but we know the final CF for a cycle takes more effort than previous
> ones, so it seems reasonable to give more time.  We have a large number
> of patches still waiting for review.

+1


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-29 Thread Robert Haas 
On Wed, Mar 29, 2017 at 3:04 PM, Alvaro Herrera
 wrote:
> Tom Lane wrote:
>> My own thought is that there's room for at least a few days' slop in
>> the end date of the final commitfest, depending on what patches remain
>> open and what the prospects are for getting them done.  (In the past
>> we've sometimes let the final fest stretch on indefinitely, which is
>> clearly the Wrong Thing; but that doesn't mean that the Right Thing is
>> to say that it ends at 2017-04-01 00:00 UTC no matter what.)  The RMT
>> should look at things in another day or two and make a judgment call
>> about that.
>
> I was rather surprised to see the March commitfest declared to exactly
> one month and feature freeze immediately thereafter.  Last time around
> we left 2 weeks between CF end and feature freeze; the previous one I
> think we had the final CF last two months.  Not stretch on indefinitely,
> but we know the final CF for a cycle takes more effort than previous
> ones, so it seems reasonable to give more time.  We have a large number
> of patches still waiting for review.

That's true, but at
https://wiki.postgresql.org/wiki/PgCon_2016_Developer_Meeting we
agreed on March 31st as the feature freeze date.  I don't think the
RMT should substantially vary a date that was agreed by a developer
meeting with wide attendance; that seems like the tyranny of the few
over the many.  To quote from that page:

===
The dates agreed were:

CF1: September 1 to 30th
CF2: November 1 to 30th
CF3: January 1 to 31st.
CF4: March 1 to 31st.
Feature Freeze: March 31st
May 16th: Beta
===

I think we should be trying to hit those dates, not re-litigate them.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-29 Thread Alvaro Herrera 
Tom Lane wrote:

> My own thought is that there's room for at least a few days' slop in
> the end date of the final commitfest, depending on what patches remain
> open and what the prospects are for getting them done.  (In the past
> we've sometimes let the final fest stretch on indefinitely, which is
> clearly the Wrong Thing; but that doesn't mean that the Right Thing is
> to say that it ends at 2017-04-01 00:00 UTC no matter what.)  The RMT
> should look at things in another day or two and make a judgment call
> about that.

I was rather surprised to see the March commitfest declared to exactly
one month and feature freeze immediately thereafter.  Last time around
we left 2 weeks between CF end and feature freeze; the previous one I
think we had the final CF last two months.  Not stretch on indefinitely,
but we know the final CF for a cycle takes more effort than previous
ones, so it seems reasonable to give more time.  We have a large number
of patches still waiting for review.

-- 
Álvaro Herrerahttps://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-29 Thread Stephen Frost 
Tom, all,

* Tom Lane (t...@sss.pgh.pa.us) wrote:
> My own thought is that there's room for at least a few days' slop in
> the end date of the final commitfest, depending on what patches remain
> open and what the prospects are for getting them done.  (In the past
> we've sometimes let the final fest stretch on indefinitely, which is
> clearly the Wrong Thing; but that doesn't mean that the Right Thing is
> to say that it ends at 2017-04-01 00:00 UTC no matter what.)  The RMT
> should look at things in another day or two and make a judgment call
> about that.

I tend to agree with this, though I am also of the opinion that it'd
be better to get a date out there earlier rather than later.  There's a
number of hackers at PGConf.US who are trying their best to review
patches and keep things moving while also giving talks and being
involved in the conference.  There's at least a couple of patches which
I can think of off-hand that I'd like to continue working through the
process to get them committed and only require another day or so of
effort, but finding that time in-between activities at the conference,
while possible, isn't ideal.

Thanks!

Stephen


signature.asc
Description: Digital signature

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-28 Thread Tom Lane 
Robert Haas  writes:
> On Thu, Mar 23, 2017 at 10:34 AM, Tom Lane  wrote:
>> As previously agreed at the PGCon 2016 developers meeting, we'll
>> institute v10 feature freeze at the completion of the current
>> commitfest (end of this month).

> This means, as I understand it, that no new features should be
> committed by anyone past the end of March, unless for some reason the
> RMT decides to grant an extension for some particular case.

Well, formally speaking, the RMT declares the onset of feature freeze.
I expressed an opinion that they'd probably do so at the end of March,
but it's their call as to exactly when to do so, and whether to grant
any limited extensions/exceptions.

My own thought is that there's room for at least a few days' slop in
the end date of the final commitfest, depending on what patches remain
open and what the prospects are for getting them done.  (In the past
we've sometimes let the final fest stretch on indefinitely, which is
clearly the Wrong Thing; but that doesn't mean that the Right Thing is
to say that it ends at 2017-04-01 00:00 UTC no matter what.)  The RMT
should look at things in another day or two and make a judgment call
about that.

> Instead,
> we should work on clearing out the open items list, so that we can go
> to beta in a timely fashion.

Certainly that should be the focus as soon as we are in feature freeze.

regards, tom lane


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

Re: [HACKERS] Schedule and Release Management Team for PG10

2017-03-28 Thread Robert Haas 
On Thu, Mar 23, 2017 at 10:34 AM, Tom Lane  wrote:
> As previously agreed at the PGCon 2016 developers meeting, we'll
> institute v10 feature freeze at the completion of the current
> commitfest (end of this month).

This means, as I understand it, that no new features should be
committed by anyone past the end of March, unless for some reason the
RMT decides to grant an extension for some particular case.  Instead,
we should work on clearing out the open items list, so that we can go
to beta in a timely fashion.

https://wiki.postgresql.org/wiki/PostgreSQL_10_Open_Items

If you have posted a bug fix patch for something committed to this
release, and it hasn't been committed yet, please a link to the thread
to the open items list.  If you have reported an issue with a v10
commit and it has not been fixed yet, please similarly add a link to
the open items list.  If, on the other hand, you committed a patch
that has generated one or more open items, please fix them.  Logical
replication is currently leading in the contest for "most unaddressed
open items" with 8, followed closely by SCRAM authentication with 6
and partitioning with 5.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers

[HACKERS] Schedule and Release Management Team for PG10

2017-03-23 Thread Tom Lane 
The Postgres release team has decided that last year's appointment
of a Release Management Team worked pretty well, so we're going to
run the version-10 release cycle the same way.  The members of the
RMT for this year will be Peter Eisentraut, Robert Haas, and Noah
Misch.

As previously agreed at the PGCon 2016 developers meeting, we'll
institute v10 feature freeze at the completion of the current
commitfest (end of this month).  The RMT will target making the
10beta1 release in mid-May and 10.0 final release in September.

Attached for reference is the RMT charter as agreed to by the
pgsql-release list.

regards, tom lane


== Proposed Release Management Team Charter

This charter takes effect on 2017-03-31 and expires three weeks after the
PostgreSQL 10 final release.

Duties:
- Meet the deadlines pgsql-release@ sets for beta1 and final release.
- When the open items list has an unowned item, promptly inform the associated
  committer that he owns that item.
- Approve the creation of REL10_STABLE, the welcoming of PostgreSQL 11
  submissions, and the resumption of CommitFests.
- Nag item owners to ensure forward progress.

Powers, exercisable by majority vote of the RMT, even in the face of no
consensus or contrary consensus:
- Approve the commit of a patch for the purpose of resolving an open item,
  including reversion commits.
- Add or remove open items.
- Enact a feature freeze.

Before using one of those powers, the Release Management Team (RMT) should
make more than zero effort to let normal community processes arrive at a
sufficient answer.  Given the aggressive schedule, the RMT may nonetheless use
these powers quickly and often.

pgsql-release@ appoints the RMT and has the power to dissolve it.

Members are Peter Eisentraut, Robert Haas, and Noah Misch.


-- 
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers