Re: [RFC 0/2] RISC-V: enable rust

[Miguel Ojeda]

On Fri, Feb 24, 2023 at 10:00 PM Conor Dooley <conor@xxxxxxxxxx> wrote:
>
> The stuff that I have lifted here all had you as the sole author in the
> "rust" branch downstream, which is why I gave you authorship. Namely:
> afba78eacb9b ("rust: generate target specification files on the fly")
> 732b3c386328 ("rust: target: remove `cpu`")
>
> I don't see anything from [1] in these commits, so I don't think that I
> made a mistake here.

It is true that I converted the original target spec files into the
script, so I added the final lines. However, in that first commit some
of the deleted files (related to RISC-V) were created by Gary. Thus it
still feels a bit wrong to not credit Gary or even mention him.

For instance, consider an even more extreme case: somebody moving a
file or doing formatting/whitespace changes. Would they be the main
and only author?

> It's RFC for a reason, I've had a poor track record with off-list emails
> to people that do not know me so would rather do it this way :)
> Probably should have noted that I wrote the ~placeholder commit messages
> though, apologies. I'll sort that out for a potential v1.

No problem!

> That's what I did! Unless I missed something that was non-obvious, the
> only name on the commits I lifted was you. Is there somewhere else I
> should have looked for that information?

I would have traced the commits back a bit more. For instance, in the
first commit you mention above, one may see the RISC-V target files
were removed, so that means something was already there. Checking who
added those files leads to a few commits from Gary (and one from
Daniel). And then it is about making a judgement call trying to be
fair :)

Cheers,
Miguel