On Fri, Dec 4, 2009 at 9:39 AM, Adam Treat <tr...@kde.org> wrote: > I don't think we should be changing the style guide for anything besides > clarifications of currently unwritten rules. No matter how the fashion may > change or how developers may change. Changing the rules throws consistency > out the window which is, I believe, the greatest benefit of having a style > guide. >
While I agree with your points re: subjectivity, and I agree that any two competent programmers can disagree on any points, it is also true that some practices can be shown to be more reliable, maintainable, or readable, and those practices do change over time, partially as technology changes and partially because this is a social process. Hence I believe that if there was a significant consensus that rules should be changed, that is okay, even if there is no semantic difference in the rule change. I agree that having style guidelines that are not followed *in the existing codebase* are pretty much useless, even if they're used for new code. But, changing the rules only throws consistency out the window if the code is not updated to conform with the rule, so an alternative to an inconsistently-formatted codebase (and a useless style guide) or style-guide stais is to require "whitespace-only" CLs to update the codebase (and yes, I know some people object to such things). Accordingly, I believe that whitespace CLs are an acceptable cost to impose as a part of this, and people proposing the change should be willing to pay the cost. Of course, I think we should attempt to determine the cost before deciding to make the change. Dunno if this changes any of your thinking or not ... Peter, were you thinking that you (or at least someone) would attempt to bring the code into compliance with the new rule? -- Dirk _______________________________________________ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
