- From: Boris Zbarsky <bzbarsky@MIT.EDU>
- Date: Wed, 08 Jun 2011 14:54:37 -0700
- To: Savil Srivastava <Savil.Srivastava@microsoft.com>
- CC: "public-web-perf@w3.org" <public-web-perf@w3.org>
On 6/8/11 2:51 PM, Savil Srivastava wrote: > As far as things go, yes, this isn't a particularly difficult to implement API expansion. However, I respectfully disagree that it doesn't have costs involved -- there are always costs of testing the extra API surface, and maintaining it over the course of the next many years. I'm assuming that since this is entirely declared in IDL the relevant code is entirely machine-generated from said idl... that's certainly the way it works in the implementations I'm familiar with. > There's also the cognitive cost of people wondering why we have two ways of doing the same thing. Yes, indeed. > In general, though, would you not agree that it is a bad principle to overload APIs without new functional benefits? Yes. -Boris
Received on Wednesday, 8 June 2011 21:55:07 UTC