Обсуждение: Non-committer reviews: is it helpful?

Hi Hackers,

I have a question for the community. While waiting for feedback on my
own patches, I was wondering: is there value in me reviewing other
people's patches, even though I am not a committer here?

I know committers are extremely busy, and I have some spare time to help
the community. I believe I can provide meaningful technical reviews and
help reduce the load on maintainers.

I’ve been around here for over a year, and sometimes it feels like the
project lacks a dedicated coordinator to direct contributor and reviewer
efforts. Because of this, I'm not always sure which patches deserve
attention first.

I saw that reviews by third parties (not contributors and committers)
are included in the commit log header.

In general, I want to understand how useful this is for committers.

Thanks!

--
Regards,
Alexander Borisov

30.01.2026 01:20, David G. Johnston пишет:
> On Thu, Jan 29, 2026 at 3:07 PM Alexander Borisov <lex.borisov@gmail.com 
> <mailto:lex.borisov@gmail.com>> wrote:
> 
>     I have a question for the community. While waiting for feedback on my
>     own patches, I was wondering: is there value in me reviewing other
>     people's patches, even though I am not a committer here?
> 
> 
> Non-committers moving patches from "needs review" to "ready for 
> committer" is a critical activity here.  It is part of the soft contract

It's just that the committers decide in the end.

> one enters into when authoring patches - to try and perform an amount of 
> review work equivalent to what you are asking from others.

Sorry, I'm not asking anything of others, I didn't quite understand your
point (sorry, English is not my native language). With my patches, I
proposed significant improvements for Postgres.

I'm just trying to understand—would a patch review from a third-party
developer be useful, would it help the committers in any way?

> As for which ones to review - everyone has their own natural 
> prioritization methodology (and capabilities) just by interacting with 
> the project and community.  I'd suggest you don't overthink things and 
> just do what feels right to you.

I am a former NGINX developer, worked with NJS (and more), and have
extensive experience in C. I just want to understand how useful reviews
from non-committers and contributors (not in list) are.
I want to be useful to the community.

I have been involved in open source for a long time, and the Postgres
community is the most mysterious to me. No one is accountable for
anything, it's anarchy, everyone decides what they need.
I'm trying to understand the rules of the "game".

--
Regards,
Alexander Borisov

On Fri, Jan 30, 2026 at 01:07:20AM +0300, Alexander Borisov wrote:
> I saw that reviews by third parties (not contributors and committers)
> are included in the commit log header.
>
> In general, I want to understand how useful this is for committers.

There are several level of reviews, of course, but I tend to find all
of them useful.  Even a small set of contributions like checking if a
patch runs or catching typo or project-style mistakes is the set of
things that helps in reducing the overall workload when a patch is
picked up to be integrated into the tree.

This is a very situational and case-by-case handling, of course.  It
makes more sense to question the design of a 3k patch than complain
about two typos in it.  A trend that I think lacks a lot in terms of
patch authors and reviews is that it is usually possible to split a
patch in multiple simpler patches, where initial pieces are more
focused on refactoring or some beautification.  I'd encourage authors
and reviewers in spending time in finding and in suggesting such
things, as it can reduce a lot the work overall.  And it's easier to
pick up for a committer.
--
Michael

Alexander Borisov <lex.borisov@gmail.com> writes:
> I have a question for the community. While waiting for feedback on my
> own patches, I was wondering: is there value in me reviewing other
> people's patches, even though I am not a committer here?

Yes, absolutely.  Even if you don't catch every problem that a
person with committer-level experience might catch, every problem
you do catch is one less thing for the eventual committer to deal
with.

Another reason why we encourage people of all experience levels
to do code reviews is that that is amazingly useful for gaining
familiarity with the Postgres code base, which is important for
becoming a more senior-level contributor.  Committers don't appear
out of nowhere; they gained the necessary knowledge by working on
patches, both their own and others'.

> I’ve been around here for over a year, and sometimes it feels like the
> project lacks a dedicated coordinator to direct contributor and reviewer
> efforts. Because of this, I'm not always sure which patches deserve
> attention first.

You presume a degree of top-down organization that doesn't exist
around here.  People work on whatever catches their fancy (or,
perhaps, what their company wants them to work on ... but that is no
business of the community at large).  That applies to reviews just as
much as to writing the patches in the first place.  So review what you
find interesting or what you think you can say something useful about.

            regards, tom lane

30.01.2026 02:41, Michael Paquier пишет:
> On Fri, Jan 30, 2026 at 01:07:20AM +0300, Alexander Borisov wrote:
>> I saw that reviews by third parties (not contributors and committers)
>> are included in the commit log header.
>>
>> In general, I want to understand how useful this is for committers.
> 
> There are several level of reviews, of course, but I tend to find all
> of them useful.  Even a small set of contributions like checking if a
> patch runs or catching typo or project-style mistakes is the set of
> things that helps in reducing the overall workload when a patch is
> picked up to be integrated into the tree.
> 
> This is a very situational and case-by-case handling, of course.  It
> makes more sense to question the design of a 3k patch than complain
> about two typos in it.  A trend that I think lacks a lot in terms of
> patch authors and reviews is that it is usually possible to split a
> patch in multiple simpler patches, where initial pieces are more
> focused on refactoring or some beautification.  I'd encourage authors
> and reviewers in spending time in finding and in suggesting such
> things, as it can reduce a lot the work overall.  And it's easier to
> pick up for a committer.

Hi Michael,

Thanks for the reply,
I'm trying to figure out how to help relieve the committers workload, as
I've seen a lot of comments on the mailing list about how busy they are.
That's where my question came from.

--
Regards,
Alexander Borisov

30.01.2026 02:44, Tom Lane пишет:
> Alexander Borisov <lex.borisov@gmail.com> writes:
>> I have a question for the community. While waiting for feedback on my
>> own patches, I was wondering: is there value in me reviewing other
>> people's patches, even though I am not a committer here?
> 
> Yes, absolutely.  Even if you don't catch every problem that a
> person with committer-level experience might catch, every problem
> you do catch is one less thing for the eventual committer to deal
> with.
> 
> Another reason why we encourage people of all experience levels
> to do code reviews is that that is amazingly useful for gaining
> familiarity with the Postgres code base, which is important for
> becoming a more senior-level contributor.  Committers don't appear
> out of nowhere; they gained the necessary knowledge by working on
> patches, both their own and others'.
> 
>> I’ve been around here for over a year, and sometimes it feels like the
>> project lacks a dedicated coordinator to direct contributor and reviewer
>> efforts. Because of this, I'm not always sure which patches deserve
>> attention first.
> 
> You presume a degree of top-down organization that doesn't exist
> around here.  People work on whatever catches their fancy (or,
> perhaps, what their company wants them to work on ... but that is no
> business of the community at large).  That applies to reviews just as
> much as to writing the patches in the first place.  So review what you
> find interesting or what you think you can say something useful about.


Thank you for your reply, it makes the community's approach clearer.
Therefore, I can confidently review patches in the mailing list and
offer my own options/edits.
With confidence that this is really useful for the community, and the
committers.

--
Regards,
Alexander Borisov