- From: Martin Thomson <martin.thomson@gmail.com>
- Date: Mon, 27 Aug 2012 12:00:11 -0700
- To: "Cullen Jennings (fluffy)" <fluffy@cisco.com>
- Cc: "public-webrtc@w3.org" <public-webrtc@w3.org>
On 27 August 2012 07:33, Cullen Jennings (fluffy) <fluffy@cisco.com> wrote: > I think you assertion that the work here is assuming the user is an idiot is just wrong and pretty distasteful. Do you have any evidence to back up this allegation? It was an essay on design, not some sort of crass allegation. To be perfectly clear, the trigger for this posting was a throwaway comment; one that did not have any significant bearing on the outcome. > ... most successful API and libraries take some complex functionality and wrap it up in a way that meets the users needs and is easy to use. API usability is important, but for the most part, usability arises from a clear communication of design intent. Part of the thesis is that architecture that supports usability is important. As requirements emerge then the architecture might need to change and with it, the design. Two concerns that I've had in this regard arise from a) the idea that certain existing structures are sacrosanct, and b) concentration on efficiency. The former leads to the sorts of hacks that we will be discussing tomorrow as it relates to ICE state. It leads to the design of new features solely to compensate for incompatibilities. The latter is an optimization. As such, it's premature. I've definitely seen an expressed desire for efficiency in this working group - though to be fair, that might have been in the context of thinking that the affected features were complete enough to justify optimization, or through some conception of design economy. --Martin
Received on Monday, 27 August 2012 19:00:39 UTC