5 ms·

The committer defends his patch, and Qu responds very constructively with this list of more important work to tackle https://lkml.org/lkml/2021/6/21/342 https:/

The committer defends his patch, and Qu responds very constructively with this list of more important work to tackle https://lkml.org/lkml/2021/6/21/342 https://lkml.org/lkml/2021/6/21/342

chapium 5y ago
Qu sounds incredibly awesome.
specialist 5y ago
Can we measure "impact" of changes? PKI measuring just commit count is pretty silly. Like SLOC. Cite folklore.org story of -10,000 lines added. Whereas Qu's suggestions are useful, important. Am totally ignorant about Linux, kernel, etc. But maybe there's already a leaderboard where the hivemind helps prioritize work.
- iicc 5y ago
  A measure of the ratio of patch effort to review effort (as subjectively judged by the reviewer, perhaps to the nearest order of magnitude) might be informative.
  - specialist 5y ago
    Belated reply. Sorry. I really like this notion. But I can't imagine what it'd look like. Please post anything you come up with. Thanks.
squiggleblaz 5y ago
That was a shockingly productive and constructive resolution. Are we doing Hacker News wrong? Should we have a system to encourage models of good and productive interchanges, which leave the participants happy and the product better? I think hostility is easier but I guess it's also addictive, even for us. I don't want to be shocked by seeing this. I want to see more of this.
Nexxxeh 5y ago
Thanks for posting this. I had initially wondered how, if not through this process, you were supposed to fix minor stuff like comments and error messages. >I'm not saying cleanup is not important, in fact we have routinely cleanups of typos/grammar for btrfs. (And I guess mostly caused by myself?) >Please at least merge all those small fixes into a larger patchset, and with a good cover letter to explain the reason (and auto-tool to do the change if possible) for all the involved maintainers, so that all of us are on the same page.

The committer defends his patch, and Qu responds very constructively with this list of more important work to tackle https://lkml.org/lkml/2021/6/21/342 https://lkml.org/lkml/2021/6/21/342

chapium 5y ago

Qu sounds incredibly awesome.

specialist 5y ago

Can we measure "impact" of changes? PKI measuring just commit count is pretty silly. Like SLOC. Cite folklore.org story of -10,000 lines added. Whereas Qu's suggestions are useful, important. Am totally ignorant about Linux, kernel, etc. But maybe there's already a leaderboard where the hivemind helps prioritize work.

iicc 5y ago
A measure of the ratio of patch effort to review effort (as subjectively judged by the reviewer, perhaps to the nearest order of magnitude) might be informative.
- specialist 5y ago
  Belated reply. Sorry. I really like this notion. But I can't imagine what it'd look like. Please post anything you come up with. Thanks.

squiggleblaz 5y ago

That was a shockingly productive and constructive resolution. Are we doing Hacker News wrong? Should we have a system to encourage models of good and productive interchanges, which leave the participants happy and the product better? I think hostility is easier but I guess it's also addictive, even for us. I don't want to be shocked by seeing this. I want to see more of this.

Nexxxeh 5y ago

Thanks for posting this. I had initially wondered how, if not through this process, you were supposed to fix minor stuff like comments and error messages. >I'm not saying cleanup is not important, in fact we have routinely cleanups of typos/grammar for btrfs. (And I guess mostly caused by myself?) >Please at least merge all those small fixes into a larger patchset, and with a good cover letter to explain the reason (and auto-tool to do the change if possible) for all the involved maintainers, so that all of us are on the same page.