Hi,
A draft of my analysis of code contributions for 2025 can be found at
https://sites.google.com/site/robertmhaas/contributions in
contributions2025-wip.dmp. In contrast to previous years, I was able
to do much more of this in an automated way this year: the principal
author of the commit was computed by grabbing the first Author or
Co-Authored-By tag from the commit message rather than by manual
inspection of all the commit messages. Yay!
Just like last year, I invite corrections from anyone who is
interested in providing them. The table of interest is commits2025,
which has columns lines and xlines. xlines is what will be used to
produce the final blog post. As usual, I've set xlines=0 if a commit
seemed to be a large, mechanical commit that shouldn't count toward
someone's lines contributed. Also, this year, for certain patches that
touched the Unicode translate tables, instead of setting xlines=0,
I've decremented it by the size of the changes to the Unicode tables,
to avoid overcounting the significance of those commits relative to
others while still giving credit for the net new code. I did not
bother to account for reverts as carefully this year, because,
thankfully, most of them touched only relatively small numbers of
lines of code, and so it didn't seem to me that they affected the
statistics very much. If I did account for reverts more carefully,
what I would do is set xlines=0 for both reverts and the reverted
commits, but only when both occurred in the same calendar year. I'm
open to feedback on whether that should be pursued further in the
interest of accuracy, but so far it didn't seem especially important
given the shape of this year's data.
My main reason for putting this out for possible corrections is to fix
author names. If the primary author of a commit is not as listed, or
where I have multiple spellings for the same person's name, or where
someone's name is not spelled as they prefer, corrections are welcome.
Secondarily, if you think I should set xlines=0 for some mechanical
commit that was not identified as such in my initial analysis, you can
also tell me about that. As before, please send corrections off-list
as proposed UPDATE statements against the commits2025 table.
Thanks,
--
Robert Haas
EDB: http://www.enterprisedb.com
