FAQ
Hi everyone,

As soon as the Perl release cycle enters a freeze, people start
committing their work in branches awaiting a merge. As blead gets
colder, hitting a full freeze, these start piling. Once 5.25.0 is out,
these are merged in, awaiting 5.25.1.

Since each development release is done once a month, and 5.25.0 was done
at the beginning of the month, it effectively means about a month and a
half for these to wait. Generally this would be fine, except it's the
first release past the freeze, meaning it's likely to collect a lot of
patches.

To accommodate this larger-than-usual set of changes, Matthew suggested
putting out 5.25.1 a bit earlier, on the 20th this month. I think that's
a good idea. What does that mean?

* On May 20th I will release 5.25.1, allowing all of these changes more
smoking time.
* Every release date stays the same, except it's the next version.
Instead of releasing 5.25.1 on June 20th, Matthew will release 5.25.2.
The stable release date stays the same as well.
* The states of freezing will all move one version forward. Instead of
"Contentious changes freeze" starting at 5.25.7, it will be start at
5.25.8. The full freeze will begin at 5.25.10 instead of 5.25.9.

If you have any objections, please let me/us know. Otherwise, 5.25.1
will be out in a week and a day on the 20th.

This is also a small reminder that blead is waiting for all your
commits. Bring it this week and you'll get longer smoking.

Thanks,
Sawyer.

Search Discussions

  • Father Chrysostomos at May 12, 2016 at 9:46 pm

    The new pumpking wrote:
    If you have any objections, please let me/us know. Otherwise, 5.25.1
    will be out in a week and a day on the 20th.
    I have none. But I do wonder why 5.25.0 was released at all, with no
    changes from 5.24 except the version number. Would anyone care to
    elaborate?
  • Ricardo Signes at May 12, 2016 at 9:49 pm
    * Father Chrysostomos [2016-05-12T17:46:47]
    The new pumpking wrote:
    If you have any objections, please let me/us know. Otherwise, 5.25.1
    will be out in a week and a day on the 20th.
    I have none. But I do wonder why 5.25.0 was released at all, with no
    changes from 5.24 except the version number. Would anyone care to
    elaborate?
    It's What We've Usually Done™

    --
    rjbs
  • Father Chrysostomos at May 13, 2016 at 4:33 am

    Ricardo Signes wrote:
    * Father Chrysostomos <sprout cpan.org> [2016-05-12T17:46:47]
    I have none. But I do wonder why 5.25.0 was released at all, with no
    changes from 5.24 except the version number. Would anyone care to
    elaborate?
    It's What We've Usually Doneâ„¢
    Not with 5.23.0, which came about three weeks after 5.22.0.
  • Abigail at May 13, 2016 at 12:09 pm

    On Thu, May 12, 2016 at 05:49:11PM -0400, Ricardo Signes wrote:
    * Father Chrysostomos [2016-05-12T17:46:47]
    The new pumpking wrote:
    If you have any objections, please let me/us know. Otherwise, 5.25.1
    will be out in a week and a day on the 20th.
    I have none. But I do wonder why 5.25.0 was released at all, with no
    changes from 5.24 except the version number. Would anyone care to
    elaborate?
    It's What We've Usually Done™

    Most of the time, the first release of a dev cycle was done shortly
    after the major release, but there have been exceptions.

    Perlhist tells use the release dates:

        Major release 1st dev release Days between
        5.12.0 Apr 12 5.13.0 Apr 20 8
        5.14.0 May 14 5.15.0 Jun 20 37
        5.16.0 May 20 5.17.0 May 26 6
        5.18.0 May 18 5.19.0 May 20 2
        5.20.0 May 27 5.21.0 May 27 0
        5.22.0 Jun 1 5.23.0 Jun 20 19
        5.24.0 May 9 5.25.0 May 9 0



    Abigail
  • Ricardo Signes at May 13, 2016 at 2:08 pm
    * Abigail [2016-05-13T08:10:39]
    On Thu, May 12, 2016 at 05:49:11PM -0400, Ricardo Signes wrote:

    It's What We've Usually Done™
    Most of the time, the first release of a dev cycle was done shortly
    after the major release, but there have been exceptions.
    Exactly why I said "usually."

    --
    rjbs
  • Aaron Crane at May 13, 2016 at 10:02 am

    Father Chrysostomos wrote:
    But I do wonder why 5.25.0 was released at all, with no
    changes from 5.24 except the version number. Would anyone care to
    elaborate?
    I've always assumed it was to make bumping the version be part of the
    task of doing the 5.even.0 stable release, rather than falling to
    whoever's doing the first real release of the new blead cycle; and
    that, in turn, that's done to ensure that the process of releasing
    blead is as consistent as possible.

    --
    Aaron Crane ** http://aaroncrane.co.uk/
  • Dave Mitchell at May 14, 2016 at 2:51 pm

    On Thu, May 12, 2016 at 09:46:47PM -0000, Father Chrysostomos wrote:
    I have none. But I do wonder why 5.25.0 was released at all, with no
    changes from 5.24 except the version number. Would anyone care to
    elaborate?
    One reason is so that 'git describe' can distinguish between a commit
    on the maint-5.24 and blead branches. So a blead commit soon after
    5.24 is released will appear as e.g. v5.25.0-XX-gYYYYYYY, while
    a maint commit will appear as v5.24.0-XX-gYYYYYY

    Otherwise they both appear confusingly as v5.24.0-something.

    --
    A major Starfleet emergency breaks out near the Enterprise, but
    fortunately some other ships in the area are able to deal with it to
    everyone's satisfaction.
         -- Things That Never Happen in "Star Trek" #13
  • Karl Williamson at May 13, 2016 at 4:19 am

    On 05/12/2016 03:31 PM, Sawyer X wrote:
    Hi everyone,

    As soon as the Perl release cycle enters a freeze, people start
    committing their work in branches awaiting a merge. As blead gets
    colder, hitting a full freeze, these start piling. Once 5.25.0 is out,
    these are merged in, awaiting 5.25.1.

    Since each development release is done once a month, and 5.25.0 was done
    at the beginning of the month, it effectively means about a month and a
    half for these to wait. Generally this would be fine, except it's the
    first release past the freeze, meaning it's likely to collect a lot of
    patches.

    To accommodate this larger-than-usual set of changes, Matthew suggested
    putting out 5.25.1 a bit earlier, on the 20th this month. I think that's
    a good idea. What does that mean?

    * On May 20th I will release 5.25.1, allowing all of these changes more
    smoking time.
    * Every release date stays the same, except it's the next version.
    Instead of releasing 5.25.1 on June 20th, Matthew will release 5.25.2.
    The stable release date stays the same as well.
    * The states of freezing will all move one version forward. Instead of
    "Contentious changes freeze" starting at 5.25.7, it will be start at
    5.25.8. The full freeze will begin at 5.25.10 instead of 5.25.9.

    If you have any objections, please let me/us know. Otherwise, 5.25.1
    will be out in a week and a day on the 20th.

    This is also a small reminder that blead is waiting for all your
    commits. Bring it this week and you'll get longer smoking.

    Thanks,
    Sawyer.
    Note that 5.25.1 is supposedly blocked from being released until the 5
    tickets still open here are fixed:

    https://rt.perl.org/Ticket/Display.html?id=127731
  • James E Keenan at May 13, 2016 at 10:27 am

    On 05/13/2016 12:18 AM, Karl Williamson wrote:
    On 05/12/2016 03:31 PM, Sawyer X wrote:
    Hi everyone,

    As soon as the Perl release cycle enters a freeze, people start
    committing their work in branches awaiting a merge. As blead gets
    colder, hitting a full freeze, these start piling. Once 5.25.0 is out,
    these are merged in, awaiting 5.25.1.

    Since each development release is done once a month, and 5.25.0 was done
    at the beginning of the month, it effectively means about a month and a
    half for these to wait. Generally this would be fine, except it's the
    first release past the freeze, meaning it's likely to collect a lot of
    patches.

    To accommodate this larger-than-usual set of changes, Matthew suggested
    putting out 5.25.1 a bit earlier, on the 20th this month. I think that's
    a good idea. What does that mean?

    * On May 20th I will release 5.25.1, allowing all of these changes more
    smoking time.
    * Every release date stays the same, except it's the next version.
    Instead of releasing 5.25.1 on June 20th, Matthew will release 5.25.2.
    The stable release date stays the same as well.
    * The states of freezing will all move one version forward. Instead of
    "Contentious changes freeze" starting at 5.25.7, it will be start at
    5.25.8. The full freeze will begin at 5.25.10 instead of 5.25.9.

    If you have any objections, please let me/us know. Otherwise, 5.25.1
    will be out in a week and a day on the 20th.

    This is also a small reminder that blead is waiting for all your
    commits. Bring it this week and you'll get longer smoking.

    Thanks,
    Sawyer.
    Note that 5.25.1 is supposedly blocked from being released until the 5
    tickets still open here are fixed:

    https://rt.perl.org/Ticket/Display.html?id=127731
    Note however that one of the items in this [META] ticket is itself a
    [META] ticket: https://rt.perl.org/Ticket/Display.html?id=123945

    I think that's a mistake. A ticket like 123945 serves merely to list
    tickets with a common theme. Such tickets are likely to remain open a
    long time (think: stack not reference-counted). They should not, IMO,
    be themselves listed as blockers for a particular release.

    If however, a particular item in the META-META ticket blocks a
    particular release, then that ticket -- and only that ticket -- should
    be listed on the release-blockage META ticket.

    So, should any of the following be listed as blockers for perl-5.25.1?

    https://rt.perl.org/Ticket/Display.html?id=36248

    https://rt.perl.org/Ticket/Display.html?id=38812

    https://rt.perl.org/Ticket/Display.html?id=85366

    You only have to look at their RT numbers to realize that these tickets
    have been open for years. Hence, they do not, prima facie, qualify as
    blockers for 5.25.1.

    If there are specific issues related to encoding.pm that *are* blockers
    to 5.25.1, *those* should be listed on 127731 -- and if they're not
    already listed on 123945, they should be.

    I'll take action along the above lines tomorrow unless someone objects
    in the mean time.

    Thank you very much.
    Jim Keenan
  • Zefram at May 13, 2016 at 10:51 am

    James E Keenan wrote:
    Note however that one of the items in this [META] ticket is itself a [META]
    ticket: https://rt.perl.org/Ticket/Display.html?id=123945
    It looks like the intent is that we'll rip out the core functionality that
    encoding.pm uses, thus making it unbuildable rather than merely faulty.
    This will resolve all outstanding encoding.pm issues at a stroke.
    In that context, it does make sense for the encoding.pm meta ticket to
    be a release blocker. If we're going that route, the ticket structure is
    fine: we'll close all the specific encoding.pm tickets and the encoding.pm
    meta ticket in one go, following the featurectomy, prior to the release.

    If we're instead going to try to fix encoding.pm, then none of its
    tickets should block 5.25.1, directly or indirectly.

    -zefram
  • Ricardo Signes at May 13, 2016 at 11:50 am
    * Karl Williamson [2016-05-13T00:18:44]
    Note that 5.25.1 is supposedly blocked from being released until the 5
    tickets still open here are fixed:

    https://rt.perl.org/Ticket/Display.html?id=127731
    My understanding of that ticket is it's really, "try to get this done before
    that release." The monthly releases are glorified snapshots, and to block one
    for a specific feature seems almost never likely to be warranted. We
    originally started using these as a way to track short-term goals, but we'd
    bump things to the next release's ticket if they didn't get done in time.

    This didn't end up being very effective, but having a "stuff to merge after
    blead opens" seemed (and was) useful. The name's a bit of a misleading name,
    though. That's on me.

    --
    rjbs

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupperl5-porters @
categoriesperl
postedMay 12, '16 at 9:31p
activeMay 14, '16 at 2:51p
posts12
users9
websiteperl.org

People

Translate

site design / logo © 2021 Grokbase