Re: Commit Fest 2019-01 is now closed

Andres Freund <andres(at)anarazel(dot)de> writes:

> There's plenty stuff that's chugging along in development but ought to
> be processed at less urgency / by different people, than the stuff
> targeted to be committed soon. It's already frustrating to contribute
> to postgresql for new people, but if they don't get feedback for half
> a year because they submitted around December / January it's almost
> guaranteed that they vanish. Additionally, there's an increasing
> amount of development projects that are too large to complete in a
> single cycle, and if we just stop looking at them for half a year,
> they'll also not succeed.

I definitely did this - and I don't think success can be declared in my
case yet. Would like to talk about that for a moment if that's alright.

GSSAPI encryption was first submitted 2015-07-02. Discussion on it
continued until 2016-08-01, when I vanished. Discussion included 118
messages, 49 of which I sent myself, and 13 separate revisions. At that
point I had gone way over the allotted time to spend on this, and had to
move on to other things. The account tracking on the commitfest app
wasn't as good then, but this corresponds to 4 commitfests.

Second push started 2018-05-23 and is ongoing. Discussion has been much
quieter - 30 messages, 10 mine - and 7 revisions (mostly due to cfbot).
Since the commitfest webpage supports github login now, the count for
second push is accurate: 5 commitfests.

So I'm at 9 commitfests total (over 2 years). The total amount of time
spent on this is incredibly daunting. And this isn't an isolated case;
for instance, at the top of the current commitfest is
https://commitfest.postgresql.org/22/528/ which has been in 14
commitfests. 14 - this'll be 3 years next month. Others have 11, 10,
10, 8... (I didn't dig into the tracking on this, so they might be
higher for the same reason my count is higher than is reflected.)

postgresql is an amazing piece of software, and it's really cool to have
something to contribute to it. And I think that the reviews I've
received have been from people who care genuinely that it keeps being
amazing. But if I weren't regularly dealing with open source and quirky
upstreams for my day job, I would be long gone. Even so, no contributor
has unlimited time; even for us corporate contributors, management will
eventually decide to solve the problem a different way.

Thanks for letting me get that off my chest. Happy hacking!

Thanks,
--Robbie