Le 22/04/2015 13:13, Gijs Kruitbosch a écrit : > On 22/04/2015 01:26, Mike Hommey wrote: >> Here are a few crude stats, gathered over the last 25271 changesets, >> assuming my pattern matching worked properly: >> >> - 1438 changesets were backed out (~5.7%) > > Thank you for some stats, which is clearly better than no stats at all! > > (Like everyone else who asked so far, I'm really curious how I'm doing!) > > However, won't this backout-changesets-rate thing be biased in favour > of people who push single csets per issue they fix, and biased against > people who fix 47-part layout refactoring bugs and land individual > changesets? :-) > > (which, you know, if they get backed out... 47 backouts right there! - > OTOH, maybe those types of csets get backed out less frequently > because they're (a) more likely to have a trypush and (b) more likely > to be tricky to back out if people land on top of it, so tough luck > and the author has to fix in place?) > > Not sure how to deal with that discrepancy between different commit > styles. Thoughts?
Not only this, but also it's not completely unexpected if big 47-parts patches get backed out for some reason. Or 47kB patches. Also, not all backouts are made to fix inbound bustage. Stats are always difficult to deal with. We _know_ they're only stats and we should not take them literally, and instead try to interprete them just like other data sources. Yet we look at them too literally. Inbound bustage should not happen. Try is not a waste of time. The time you wait for your build result... well you can do something else during that time, right? If you're really confident your work won't break anything then you'll likely only wait a few hours before landing, it's not a big deal. And if the try run catches an issue, you saved yourself and sheriffs and people bisecting and everybody else a lot of time. Just use try. -- Julien
signature.asc
Description: OpenPGP digital signature
_______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
