As I was trying to decide who is the author of the ob-java docs, I realized
it's not clear how you're defining authorship due to my confusion about
ob-java.  I can think of three ways to determine authorship:

1. the person that wrote it
2. the people who influenced the code
3. the first person to check in the filename

At first I thought I wrote ob-java by rule 1.  I didn't start from the old
ob-java, and I replaced the entire file.  The patch shows only 10 random
lines of over 400 matched the original ob-java.  If we don't count the
lines that also match ob-template.el, there are only 5.

When you said I didn't write it I thought rule 2 was the next most
reasonable, so I made the authors those that wrote the code that I
referenced.  But after thinking about it more it can't be this.  Adding
languages to babel isn't documented well enough for anyone to do it without
looking at an existing implementation, so going by rule 2 all languages
would be authored by whoever wrote the first one, and they're not.

I'm not sure but I think you'd say I wrote ob-haxe, the ob-haxe tests, the
ob-java tests, and the ob-java docs, but not ob-java.  These match up with
rule 3.  I don't think rule 3 is the one anyone would pick from the list,
but maybe most would subconsciously use it as a heuristic for rule 1, since
rule 1 is hard to establish.  I think the change in authorship is clear for
ob-java because it was replaced in one patch, usually changes are
incremental.  Each file is The Ship of Theseus.  Even if we took the
trouble to determine how much any person wrote, it is difficult to decide
for oneself let alone agree on the amount of change required to establish
new authorship.

But rule 3 doesn't work if a file is rewritten.  If Dostoevsky checks in
the text of "Crime and Punishment" as book.txt, and then Dr. Seuss replaces
the content with "The Cat in the Hat," we'd have to say Dostoevsky wrote
"The Cat in the Hat."

So I think either you didn't notice that I'd replaced the file, or you
considered the lines that matched sufficient for continuity, or you're
thinking about authorship in a way I haven't guessed.  Could you clarify?

On Wed, Oct 28, 2020 at 5:13 AM Bastien <bzg@gnu.org> wrote:

> ian martins <ianxm@jhu.edu> writes:
>
> > But I want to follow your conventions. I will put the authors of ob-C
> > and ob-python (Eric Schulte and Dan Davison) as the authors of
> > ob-java and ob-haxe. The implementations are nearly the same. it
> > wouldn't make sense for them to have different authors.
>
> Thanks for doing so!
>
> --
>  Bastien
>