On 10/7/25 1:35 PM, Eric Rescorla wrote:
On Tue, Oct 7, 2025 at 11:22 AM Eliot Lear <[email protected]> wrote:
Hi Eric,
On 07.10.2025 20:09, Eric Rescorla wrote:
I'm struggling to understand the concern here, given that the
current approved email flow involves the editors sending
text in OLD/NEW format, which is essentially what GitHub
suggestions are. There's no ambiguity about the desired
new state and if they are committed it's trivial to see what
changed.
That's not ENTIRELY the case (so to speak). Sometimes there are
some global changes with cassification, spelling and such,
requiring rebasing, etc.
Are you talking about cases in which the RPC made some change
that the author wants reversed globally?
No, I think he's talking about places where the authors are asking for a
change in lots of places by writing one sentence in the current auth48
response rather than using and OLD/NEW for each instance.
Yes, I agree that in that
case you wouldn't use suggestions. But you also wouldn't attach
a PR update, you would tell the RPC in text to reverse it.
I don't see why this would require a rebase, though.
I'm *fully confident* the RPC is up to managing that, btw, in
terms of finding the best work flow (like maybe ordering PR
processing). And I don't want to make a mountain out of a mole
hill, but there will probably be times when ```suggestion doesn't
QUITE cut it, and we shouldn't rely on it as the only means for
changes.
Nor am I suggesting that.
-Ekr
Regards,
Eliot
_______________________________________________
rfc-interest mailing list [email protected]
To unsubscribe send an email [email protected]
_______________________________________________
rfc-interest mailing list -- [email protected]
To unsubscribe send an email to [email protected]
