Jakub Kicinski <kuba@xxxxxxxxxx> writes: > On Mon, 13 Dec 2021 16:48:37 +0100 Toke Høiland-Jørgensen wrote: >> > Status notification when checks are failing? Hopefully not, we don't >> > want people posting patches just to get them tested... >> >> Well no, but sometimes a patch will have failures despite the best >> efforts of the submitter (otherwise what's the point of the checks?). >> Right now the only way for me to discover that there's an issue is to go >> look at the patchwork web interface, and I wanted something that better >> suits my workflow (i.e., that's not in a web browser). > > I think that the maintainer should notify the submitter about > the reason the patch state was changed (with the exception of > patches for a different tree, maybe). I know Kees has been > trying to add more meaningful states to patchwork but I can > never guess the meaning of those either :S So no automated > state checker can replace the maintainer's reply. Well, sometimes the maintainers forget to reply entirely :) And yeah, I do realise that no bot is going to be able to tell me exactly what the current status is, I just want a tool to help me manually keep track... >> I wasn't asking for patchwork to send out automatic notifications >> (yikes!), I just wanted to know if anyone else had done something >> similar before I go play around with the patchwork API myself... :) > > Despite the promise of "best effort" I fear such automation. > It's pretty common in (let's call them) modern workflows to > submit PRs / post changes just to get them tested by a CI. > We don't want to give people the impression that the mailing > list can serve this purpose. People could already do that, though, they just have to look at the patchwork site. And I really don't think it's that convenient of an interface that this is a real risk. But don't worry, I'm not planning to publish this as a generally-available service :) -Toke