ignite-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Anton Vinogradov ...@apache.org>
Subject Re: Clean up of our PRs and IEPs before 2019
Date Thu, 25 Jul 2019 10:44:02 GMT
Folks,

Is it possible just to ignore obsolete PRs somehow?
Not sure this crusade worth it.

On Thu, Jul 25, 2019 at 1:18 PM Павлухин Иван <vololo100@gmail.com> wrote:

> Maxim,
>
> Quite a nice idea. Could we go even further? Add a comment to each 1-2
> year old PR asking if the author could close it (most likely with help
> of some automation). As I know GitHub sends emails with PR comments to
> authors.
>
> чт, 25 июл. 2019 г. в 13:05, Dmitriy Pavlov <dpavlov@apache.org>:
> >
> > Folks, please close not needed PRs.
> >
> > I don't have contact with Pyatkov & dkarachentsev. Folks, please step in.
> > Also, feel free to reopen PRs if you still want change to be merged.
> >
> > чт, 25 июл. 2019 г. в 12:39, Maxim Muzafarov <maxmuzaf@gmail.com>:
> >
> > > Folks,
> > >
> > > Can we contact with some members manually and ask them to close unused
> > > PRs? Most of the users are active community members, so I think they
> > > will respond quite fast.
> > >
> > > I've briefly checked GitHub:
> > >
> > > dkarachentsev - 62 opened PRs
> > > ilantukh - 58 opened PRs
> > > dgovorukhin - 44 opened PRs
> > > mcherkasov - 23 opened PRs
> > > ascherbakoff  - 22 opened PRs
> > > vldpyatkov - 21 opened PRs
> > >
> > > On Thu, 25 Jul 2019 at 12:28, Dmitriy Pavlov <dpavlov@apache.org>
> wrote:
> > > >
> > > > Hi Alexey,
> > > >
> > > > second need it to check all open PRs from community members for
> fixes,
> > > > which could be merged to Ignite codebase.
> > > >
> > > > Which is why I'm not so sure that we should automatically close. I
> ask
> > > > everyone to close their PRs, and I manually double-check PRs remained
> > > > opened.
> > > >
> > > > The third need is to automatically tests all opened PRs and provide
> visas
> > > > to every PR we have. In case we have PRs with 0 blockers we should
> take
> > > it
> > > > into review process. No all newcomers aware of TC Bot, so I would
> like to
> > > > automate this process as much as possible.
> > > >
> > > > Sincerely,
> > > > Dmitriy Pavlov
> > > >
> > > > чт, 25 июл. 2019 г. в 12:22, Alexey Zinoviev <zaleslaw.sin@gmail.com
> >:
> > > >
> > > > > The long period totally reduces the discontent and outrage of
> community
> > > > > members (if you reduce to 2-6 weeks it could be intersected with
> human
> > > > > events of most part of contributors like vacation, birthdays,
> wedding,
> > > spam
> > > > > filters and etc.), believe me (I have the same experience as I
> > > mentioned)
> > > > >
> > > > > From the other hand, what the real reason to reduce it to the
> shorter
> > > > > period? Bot needs? Robot needs?
> > > > > Robot could wait, I hope:)
> > > > >
> > > > > чт, 25 июл. 2019 г. в 14:08, Павлухин Иван <vololo100@gmail.com>:
> > > > >
> > > > > > Alexey,
> > > > > >
> > > > > > Yep, I imagined a similar procedure in my mind. Just curious,
> why do
> > > > > > you think that a period before actions are taken should be so
> long
> > > > > > (3-6 months)?.
> > > > > >
> > > > > > чт, 25 июл. 2019 г. в 11:55, Alexey Zinoviev <
> zaleslaw.sin@gmail.com
> > > >:
> > > > > > >
> > > > > > > Dear Igniters, I have one suggestion
> > > > > > >
> > > > > > > If a most of commiters will support idea of automatic
> "cleaning",
> > > we
> > > > > > should
> > > > > > > provide next options
> > > > > > >
> > > > > > >    - declare a long period for putting labels or leaving
> comments
> > > for
> > > > > > >    useful PRs from their authors (about 3-6 months)
> > > > > > >    - generate notifications for all authors of all PRs
with
> > > > > clarification
> > > > > > >    of our goals
> > > > > > >    - every month reminder in dev-list and via e-mail to
each
> PR's
> > > > > author
> > > > > > >
> > > > > > > The best way, of course, the closing by our hands in each
> module
> > > and
> > > > > area
> > > > > > > with tags "obsolete" or something else.
> > > > > > >
> > > > > > > P.S. I was in the same situation in Open Street Map community
> and
> > > the
> > > > > > > principles for automated cleaning were the same like suggested
> by
> > > > > myself
> > > > > > > above
> > > > > > >
> > > > > > > I hope that we will be careful with our community
> > > > > > >
> > > > > > > чт, 25 июл. 2019 г. в 13:23, Dmitriy Pavlov <
> dpavlov@apache.org>:
> > > > > > >
> > > > > > > > Nikolay, committer could after setting up a link between
GH &
> > > Apache
> > > > > > > > accounts.
> > > > > > > > https://gitbox.apache.org/setup/
> > > > > > > >
> > > > > > > > чт, 25 июл. 2019 г. в 11:17, Nikolay Izhikov
<
> > > nizhikov@apache.org>:
> > > > > > > >
> > > > > > > > > Yes.
> > > > > > > > >
> > > > > > > > > Do someone have permission to close my(or any
other
> > > contributor) PR
> > > > > > to
> > > > > > > > > apache/ignite?
> > > > > > > > >
> > > > > > > > > В Чт, 25/07/2019 в 11:05 +0300, Павлухин
Иван пишет:
> > > > > > > > > > NIkolay,
> > > > > > > > > >
> > > > > > > > > > Do you mean technical ability?
> > > > > > > > > >
> > > > > > > > > > чт, 25 июл. 2019 г. в 10:33, Nikolay
Izhikov <
> > > > > nizhikov@apache.org
> > > > > > >:
> > > > > > > > > > >
> > > > > > > > > > > Hello, Ivan.
> > > > > > > > > > >
> > > > > > > > > > > Do we have the ability to close PRs
from other
> > > contributors?
> > > > > > > > > > >
> > > > > > > > > > > В Чт, 25/07/2019 в 09:12 +0300,
Павлухин Иван пишет:
> > > > > > > > > > > > Igniters,
> > > > > > > > > > > >
> > > > > > > > > > > >  I would like to resume a discussion
about PRs
> cleanup.
> > > > > > > > Additionally
> > > > > > > > > > > > to concerns provided earlier some
TC Bot operations
> are
> > > > > slowed
> > > > > > down
> > > > > > > > > > > > due to a huge amount of open PRs.
> > > > > > > > > > > >
> > > > > > > > > > > > As time has passed, I ask you
all again to share an
> > > opinion
> > > > > > about
> > > > > > > > > > > > centralized cleanup of obsolete
PRs. Also, a precise
> > > criteria
> > > > > > to
> > > > > > > > > > > > consider PR as obsolete is a subject
for dicsussion
> as
> > > well.
> > > > > > > > > > > >
> > > > > > > > > > > > чт, 13 дек. 2018 г. в 11:55,
Petr Ivanov <
> > > > > mr.weider@gmail.com
> > > > > > >:
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > > On 11 Dec 2018, at 10:10,
Nikolay Izhikov <
> > > > > > nizhikov@apache.org
> > > > > > > > >
> > > > > > > > > wrote:
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Hello, Ivan.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Personally, I keep my
PR's clear.
> > > > > > > > > > > > > > So, I don't have dozens
of opened PR.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > But, I don't support
Dmitriy proposal for several
> > > > > reasons:
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > 1. We introduce some
new, not required, level of
> > > > > > bureaucracy.
> > > > > > > > > > > > > > From my experience -
not required bureaucracy is
> a
> > > BAD
> > > > > > thing.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > 2. We spread our work
pattern to whole community.
> > > > > > > > > > > > > > I believe there are
many patterns of dealing with
> > > *YOUR
> > > > > > OWN*
> > > > > > > > PRs.
> > > > > > > > > > > > > > Some of them can lead
to dozens of opened PRs to
> > > master.
> > > > > > > > > > > > > > Whats wrong with it?
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > 3. I dont' see any issues
with many opened PRs.
> > > > > > > > > > > > > > What problem we trying
to solve?
> > > > > > > > > > > > >
> > > > > > > > > > > > > But I see.
> > > > > > > > > > > > > Lots of opened PRs (and obsolete
branches as well)
> > > consumes
> > > > > > huge
> > > > > > > > > amount of data and time when TC performs changes
detect
> > > operations
> > > > > > (every
> > > > > > > > > minute, BTW).
> > > > > > > > > > > > > Also, IMO, ORDER is not an
unnecessary level of
> > > > > bureaucracy,
> > > > > > but
> > > > > > > > > part of the project development workflow in area
of
> cleaning
> > > up and
> > > > > > > > keeping
> > > > > > > > > everything fresh and actual.
> > > > > > > > > > > > >
> > > > > > > > > > > > >
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > 4. Closing abanodned
PRs doesn't force anybody to
> > > review
> > > > > > the
> > > > > > > > > rest.
> > > > > > > > > > > > > > Instead of ordering
something to one way or
> another,
> > > > > let's
> > > > > > > > solve
> > > > > > > > > real problem:
> > > > > > > > > > > > > >
> > > > > > > > > > > > > >       - help the community
doing PR review.
> > > > > > > > > > > > > >       - fixing failing
tests.
> > > > > > > > > > > > > >       - introducing
new code inspections to make
> our
> > > code
> > > > > > base
> > > > > > > > > clear.
> > > > > > > > > > > > > >       - making Ignite
improvements
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > 5. I don't see how our
numbers differs from other
> > > Apache
> > > > > > > > projects
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > Apache Kafka - 533 PR
opened.
> > > > > > > > > > > > > > Apache Spark - 484 PR
opened.
> > > > > > > > > > > > > > Apache Flink - 430 PR
opened.
> > > > > > > > > > > > > >
> > > > > > > > > > > > > > В Вт, 11/12/2018
в 09:24 +0300, Pavel Tupitsyn
> пишет:
> > > > > > > > > > > > > > > Agree with Dmitriy.
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > We use GitHub PRs
in our workflow, therefore we
> > > should
> > > > > > keep
> > > > > > > > > them in order.
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > We can close PRs
that refer to closed tickets,
> > > this can
> > > > > > be
> > > > > > > > > done with a
> > > > > > > > > > > > > > > simple script.
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > On Tue, Dec 11,
2018 at 9:15 AM Павлухин Иван <
> > > > > > > > > vololo100@gmail.com> wrote:
> > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > Nikolay,
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > I must say
that when I first saw 1K+ open
> PRs my
> > > > > first
> > > > > > > > > thought was
> > > > > > > > > > > > > > > > that something
was wrong with a review
> process.
> > > In my
> > > > > > mind
> > > > > > > > > in not very
> > > > > > > > > > > > > > > > big project
open PR list can reflect very
> well
> > > the
> > > > > real
> > > > > > > > work
> > > > > > > > > in
> > > > > > > > > > > > > > > > progress.
For bigger projects things become
> more
> > > > > > > > complicated.
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > Dmitriy,
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > Do you have
some cleanup automation in mind?
> > > > > > Immediately I
> > > > > > > > > think that
> > > > > > > > > > > > > > > > it is fully
safe to close all PRs that were
> not
> > > > > touched
> > > > > > > > more
> > > > > > > > > than a
> > > > > > > > > > > > > > > > year.
> > > > > > > > > > > > > > > > пн, 10 дек.
2018 г. в 20:01, Dmitriy Pavlov <
> > > > > > > > > dpavlov@apache.org>:
> > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > The main
concern is related to chances that
> > > > > newcomer
> > > > > > will
> > > > > > > > > have to obtain
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > a
> > > > > > > > > > > > > > > > > review
support from the community.
> > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > Actually,
a lot of people doing their best
> to
> > > > > > provide a
> > > > > > > > > feedback to
> > > > > > > > > > > > > > > > > newcomers,
and count of issues still in PA
> > > state
> > > > > goes
> > > > > > > > down
> > > > > > > > > (84 is a
> > > > > > > > > > > > > > > > > relatively
small count of issues in PA
> state).
> > > But
> > > > > > 1428
> > > > > > > > > PRs may imply we
> > > > > > > > > > > > > > > > > don't
review here, as we have tons of
> > > incomplete
> > > > > PRs.
> > > > > > > > > Actually, most of
> > > > > > > > > > > > > > > > > these
PRs were merged (but not using
> > > > > > > > > ./apply-pull-request.sh script, but
> > > > > > > > > > > > > > > > > manually,
without reference to PRs).
> > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > Another
benefit of revising this list, if
> > > there are
> > > > > > any
> > > > > > > > > changes which
> > > > > > > > > > > > > > > > > were
not accomplished with a proper ticket
> > > with PA
> > > > > > > > status,
> > > > > > > > > we will
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > identify
> > > > > > > > > > > > > > > > > a number
of additional contributions to be
> > > applied
> > > > > > to the
> > > > > > > > > codebase.
> > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > пн,
10 дек. 2018 г. в 19:53, Nikolay
> Izhikov <
> > > > > > > > > nizhikov@apache.org>:
> > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > >
Hello, Dmitriy.
> > > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > >
What, exactly concerns newcomers?
> > > > > > > > > > > > > > > > > >
What is wrong with opened PR?
> > > > > > > > > > > > > > > > > >
How project will benefit from closed PR?
> > > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > >
> The same proposal is related to IEP
> > > statuses.
> > > > > If
> > > > > > you
> > > > > > > > > were involved
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > in an
> > > > > > > > > > > > > > > > > >
IEP, please validate its status
> > > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > >
+1. We should maintain IEP description
> up to
> > > > > date.
> > > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > >
пн, 10 дек. 2018 г., 19:15 Dmitriy Pavlov
> > > > > > > > > dpavlov@apache.org:
> > > > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > > >
> Hi Igniters,
> > > > > > > > > > > > > > > > > >
>
> > > > > > > > > > > > > > > > > >
> Newcomers to Apache Ignite sometimes
> became
> > > > > > concerned
> > > > > > > > > about many
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > open PRs
> > > > > > > > > > > > > > > > > >
> in the project. Apache Ignite TC Bot
> also
> > > > > > performs
> > > > > > > > > runs checks with
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > a PR
> > > > > > > > > > > > > > > > > >
> open. Apache Ignite pulls list
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > https://github.com/apache/ignite/pulls
> > > > > > > > > > > > > > > > > >
> contains
> > > > > > > > > > > > > > > > > >
> 1442 PRs open while only 84 issues are
> > > waiting
> > > > > > for
> > > > > > > > > review.
> > > > > > > > > > > > > > > > > >
>
> > > > > > > > > > > > > > > > > >
> Could you please verify the list of
> your
> > > PRs in
> > > > > > > > Apache
> > > > > > > > > Ignite
> > > > > > > > > > > > > > > > > >
> https://github.com/pulls  and close
> every
> > > not
> > > > > > > > > needed/already merged
> > > > > > > > > > > > > > > > > >
> change?
> > > > > > > > > > > > > > > > > >
>
> > > > > > > > > > > > > > > > > >
> The same proposal is related to IEP
> > > statuses.
> > > > > If
> > > > > > you
> > > > > > > > > were involved
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > in an
> > > > > > > > > > > > > > > > > >
> IEP, please validate its status here
> > > > > > > > > > > > > > > > > >
>
> > > > > > > > >
> > > > >
> https://cwiki.apache.org/confluence/display/IGNITE/Active+Proposals
> > > > > > > > > > > > > > > > > >
> Please
> > > > > > > > > > > > > > > > > >
> set a correct state for your IEP, as
> well.
> > > > > > > > > > > > > > > > > >
>
> > > > > > > > > > > > > > > > > >
> Should you have any questions, please
> don't
> > > > > > hesitate
> > > > > > > > > to ask here.
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > Thank
> > > > > > > > > > > > > > > > > >
you
> > > > > > > > > > > > > > > > > >
> in advance!
> > > > > > > > > > > > > > > > > >
>
> > > > > > > > > > > > > > > > > >
> Sincerely,
> > > > > > > > > > > > > > > > > >
> Dmitriy Pavlov
> > > > > > > > > > > > > > > > > >
>
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > >
> > > > > > > > > > > > > > > > --
> > > > > > > > > > > > > > > > Best regards,
> > > > > > > > > > > > > > > > Ivan Pavlukhin
> > > > > > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > --
> > > > > > Best regards,
> > > > > > Ivan Pavlukhin
> > > > > >
> > > > >
> > >
>
>
>
> --
> Best regards,
> Ivan Pavlukhin
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message