PostgreSQL development has certainly professionalized over the years, in the sense that more people do it as a job now rather than (only) as a hobby. I was wondering whether that would be visible in the time of day when people do publicly visible PostgreSQL work.

We can get the time of all commits over time:

git log --format='%ci'

Cut off the time zone (We want the local time of the committer) and save it to a file:

git log --format='%ci' | cut -d ' ' -f 1,2 > commit-times.txt

Import into PostgreSQL:

create table commit_times (t timestamp);

(Note I’m using the dreaded timestamp without time zone type here!)

\copy commit_times from commit-times.txt

Now let’s look at the most common hour things were committed each year:

select extract(year from t) as year,
       count(*),
       mode() within group (order by extract(hour from t))
from commit_times
group by extract(year from t)
order by 1;

 year | count | mode
------+-------+------
 1996 |   876 |    6
 1997 |  1698 |    3
 1998 |  1744 |    4
 1999 |  1788 |    2
 2000 |  2535 |   18
 2001 |  3061 |   19
 2002 |  2654 |   20
 2003 |  2416 |   22
 2004 |  2548 |   22
 2005 |  2419 |   21
 2006 |  2152 |   22
 2007 |  2188 |   21
 2008 |  1652 |   19
 2009 |  1388 |   20
 2010 |  1801 |   16
 2011 |  2029 |   12
 2012 |  1604 |   22
 2013 |  1369 |   12
 2014 |  1744 |   15
 2015 |  1816 |   13
 2016 |  2073 |   12
 2017 |  2477 |   11
 2018 |  2127 |   11
 2019 |  2130 |   11
 2020 |  2181 |   13
 2021 |  2271 |   11
 2022 |  2358 |   11

You can clearly see this changing over the years. In the first few years, people were apparently working under the cover of darkness?!? Then, from about 2000 to 2010, most commit work happened after office hours. Since then, most commit work has clearly been during the day time.

I can also look back at my own commits:

 year | count | mode
------+-------+------
 2000 |   331 |   16
 2001 |   396 |   22
 2002 |   166 |   18
 2003 |   238 |   22
 2004 |   169 |   21
 2005 |    67 |    9
 2006 |    88 |   11
 2007 |   133 |   14
 2008 |   140 |   14
 2009 |   249 |   21
 2010 |   134 |   20
 2011 |   252 |   23
 2012 |   266 |   20
 2013 |   198 |   22
 2014 |   140 |   21
 2015 |   123 |   21
 2016 |   214 |   12
 2017 |   561 |   11
 2018 |   330 |   11
 2019 |   301 |    9
 2020 |   261 |    8
 2021 |   249 |    8
 2022 |   302 |    9

This makes sense: Around 2005–2008 and since 2016, I have been working for PostgreSQL services companies, so a lot of the community work has been happening during the day. The other times, I was not, so I was mostly active “at night”. It’s also quite visible that after I moved from the U.S. to Europe in 2018, my activity moved to earlier in the day. That’s because mornings are the quieter part of the work day now (fewer meetings, fewer emails), but I also recall that I made a conscious shift to not commit things late in the day anymore, after having gotten burned by the build farm a few times.