On Thu Jan 22 2015 at 1:10:27 PM Bryan Davis <bd...@wikimedia.org> wrote:
> On Thursday, January 22, 2015, Mark A. Hershberger <m...@nichework.com> > wrote: > > > Legoktm <legoktm.wikipe...@gmail.com <javascript:;>> writes: > > > > > On 01/21/2015 09:39 AM, Jeroen De Dauw wrote: > > >> Hey, > > >> > > >> Does the new syntax offer any advantage over the old one? > > > > > > It's a little bit faster by cutting down one function call which adds > up > > > when a lot of hooks are called. > > > > "adds up" is a poor defense for creating work for end users and > > developers. Has anyone actually measured what the difference is or is > > this just an example of premature optimization[1]? > > > > Mark. > > > > Footnotes: > > [1] http://en.wikipedia.org/wiki/Program_optimization#When_to_optimize > > > > I don't have before/after measurements but Aaron started changing from > wfRunHooks to direct Hook calls based on data gathered from WMF production > profiling. There was a non-trivial performance impact of the additional > layer of indirection. > > Sure, but obvious performance gains are obvious. The real question to me is "was moving it into a static function inside of Hooks a benefit over it just being a global function?" My guess is no, not really. -Chad _______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
