10 Tips for Better Pull Requests

[ketmar via Digitalmars-d]

On Fri, 16 Jan 2015 17:16:38 +0000
Tobias Pankrath via Digitalmars-d <digitalmars-d at puremagic.com> wrote:

> On Friday, 16 January 2015 at 16:22:13 UTC, ketmar via 
> Digitalmars-d wrote:
> > On Fri, 16 Jan 2015 08:10:50 -0800
> > Andrei Alexandrescu via Digitalmars-d 
> > <digitalmars-d at puremagic.com>
> > wrote:
> >
> >> On 1/16/15 7:50 AM, H. S. Teoh via Digitalmars-d wrote:
> >> > On Thu, Jan 15, 2015 at 08:20:32PM -0800, Walter Bright via 
> >> > Digitalmars-d wrote:
> >> >> http://blog.ploeh.dk/2015/01/15/10-tips-for-better-pull-requests/
> >> >>
> >> >> I agree with pretty much everything in this article.
> >> >>
> >> >> tl,dr:
> >> >>
> >> >> "The more you make your reviewer work, the greater the risk 
> >> >> is that
> >> >> your Pull Request will be rejected."
> >> >
> >> > In the case of D, "the more you make your reviewer(s) work, 
> >> > the greater
> >> > the risk is that your Pull Request will sit in the queue 
> >> > forever." :-P
> >> 
> >> I think it would be great if we defined a simple policy for 
> >> closing pull requests that are lingering. -- Andrei
> > it sits in queue without any comments more than 20 days? reject 
> > and
> > close it.
> 
> Bad idea. Take for example this one of mine 
> https://github.com/D-Programming-Language/phobos/pull/2793 that 
> sits there for more than 20 days. I've addressed all concerns and 
> now it's waiting for someone who feels responsible for 
> std.container to pull it.
i still believe that this is a great idea. either it is significant
enough to get some traction, or it doesn't needed. if upstream cannot
upkeep with amount of changes, it's better to decreare incoming volume.
"close after 20 days" is a "traffic shaper": if nobody cares for
something withing 20 days (and nobody cares enough even to write "i'll
busy now, but i plan to take a look at this withing next 10 days"),
there is no sense to leave it rotting. what leaving it open does is
stealing the original author's time which he dedicates to keep a PR "in
shape" just in case that after another year somebody will look at it
again.

so what devteam does with keeping it open is stealing other's time. and
if enough time will pass withour author busy keeping PR up-to-date, it
will eventually be closed as "unmergeable" anyway. 20 days of
inactivity is enough to mark the thing as "not required right now".

> 4. Someone says, that the pull is rejected, because it was 
> sitting there for more than 20 days. It would be my very last 
> pull request. Period.
so you believe that it better left rotting without any sign of life for
monthes and only then someone will close it? so you'd better spend some
time in a hope that "maybe something will change" instead of clearly
see that nobody has enough time to look at it now, so it's better
be closed before it rots and stop eating your attention? nobody tells
that you cannot make a new PR with the same content if you really feels
that something must be done, along with NG posts.

the fact is that 20 days of inactivity is more than enough to tell
that PR can be safely closed, as reviewers have more important things
to do. they *always* will have more important things to do, so PR will
be resheduled and resheduled and then it will rot and became inedible.
or someone will stumble upon it on accident and yahoo! after seven and
a half monthes it will be accepted... maybe.

see -- the best way to make someone react to your PR is to spam NG with
requests to review it. or wait indefinitely.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: not available
URL: <http://lists.puremagic.com/pipermail/digitalmars-d/attachments/20150116/873f2aec/attachment.sig>