On 20/12/13 14:35, Jeff Gilbert wrote: > How do patches get fully-reviewed if the sum-knowledge of the > reviewers doesn't include whether the style's right? Alternatively, > who is signing off on the content of code, having in-depth knowledge > of how it works, but not knowing the style? That sounds like a > different problem, really. (I've also definitely seen this, fwiw. A > number of times, I've seen mis-styled code that should have failed > review for other reasons. Not that I'm saying this is a usecase of > differing styles! Rather, mis-styling correlates with poor review > quality)
Disinterest in pedantic white space doesn't undermine someone's ability to assess code structure. It's the characters in between the white space that matter. > If we're going to standardize a certain style for the whole tree, > then we're going to need to re-discuss a couple of the rules. It's a > decent amount of work to restyle the modules well, and I'm hardly > eager to do work to make this code *less* readable for the only > people who ever look at it. I'm personally fine with bikeshedding > over this, but I'm not going to jump on the 'standardize' bandwagon > if we're just going to be told not to bikeshed later when > counter-proposals are brought up. There needs to be compromise one > some front. How about we just standardise on clang-format. Patches accepted. You can format the lines you've changed with this command: $ hg diff -U0 -r tip^ | clang-format-diff-3.4 -p1 -style=Mozilla I've added a patch for mach clang-format into bug 952379. The Mozilla style could do with some work. > Besides, if style is important enough to standardize, surely it must > be important enough to fully address. That doesn't follow. Consistency is valuable. Everything else is preference. Anthony _______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
