- From: Erik Wilde <dret@berkeley.edu>
- Date: Mon, 16 Mar 2015 17:35:19 +0100
- To: James M Snell <jasnell@gmail.com>, "public-socialweb@w3.org" <public-socialweb@w3.org>
hello all. On 2015-03-16 17:24, James M Snell wrote: > All, one of the items we're going to need to move forward on Activity > Streams 2.0 completion is a Test Suite. Unfortunately I do not have > the free cycles at the moment to work on it myself. We need someone to > step up and volunteer to begin working on this item. What I imagine, > at this point, is that the test will essentially amount to a validator > -- validating that the output generated by an implementation is > correct and validating that the implementation can properly roundtrip > correct input or fail appropriately on incorrect input. also, it would be great if the test suite would not just say when to fail, but also say what to report to applications when encountering representations containing non-standard vocabulary content. this would be the litmus test for openness and extensibility: what kind of inputs are accepted (and which ones are discarded), are implementations expected to report those to applications, and in which particular way. to me, this is very similar to an infoset: we need to say clearly what does and doesn't matter, in abstract AS terms. for example, property sequence in an AS object (probably) doesn't matter (and thus does not need to be reported to applications), but properties themselves always do matter, even if implementations happen to not have built-in support for a particular property name. cheers, dret. -- erik wilde | mailto:dret@berkeley.edu - tel:+1-510-2061079 | | UC Berkeley - School of Information (ISchool) | | http://dret.net/netdret http://twitter.com/dret |
Received on Monday, 16 March 2015 16:35:50 UTC