- From: Marcos Caceres <w3c@marcosc.com>
- Date: Mon, 14 Jan 2013 20:24:37 +0000
- To: Clint Hill <clint.hill@gmail.com>
- Cc: Brian Kardell <bkardell@gmail.com>, "public-nextweb@w3.org" <public-nextweb@w3.org>
Ok, I've sent a pull request with tests for Byte, Object, Boolean. Clint's test format worked a treat! :D https://github.com/extensibleweb/webidl.js/pull/36 Clint, I didn't remove your boolean tests, but we probably don't need them. Also, it is cool if we use WebIDL.Whatever-tests.js as the naming convention? It's a bit shorter and bit easier for me to read. -- Marcos Caceres On Monday, January 14, 2013 at 6:29 PM, Marcos Caceres wrote: > Hi Clint, > > On Monday, January 14, 2013 at 6:10 PM, Clint Hill wrote: > > > For the moment - disregard the WebIDL interface tests - I'd like to > > rewrite those to make them use the require statement as well (no need to > > use setTimeout the way I did). > > > > Ok, no probs. > > Modules don't switch (not sure what you mean by that exactly). > > > > Just the order in which they are loaded. > > RequireJS > > holds an identity map of modules and when the test modules are loaded they > > could make use of the entire set of modules that are loaded by the > > original require func (as done by require.js with data-main). > > > > Ok, as I had changed that from the original to a new file I confused myself :) > > Qunit starts > > on any test func call. > > > > Ok. I guess this answers my question. I'd stopped auto running tests by disabling it in my own configuration (QUnit.config.autostart = false;). I'm starting to see how it works now. > > Load order actually could be guaranteed within the > > dependencies array of the require func if we were to include the order > > plugin. But as for each module, that load order is based on the script > > include order in index.html. > > > > Thanks Clint, that explanation was really helpful! :) I'll finish up the Byte tests and then swap over to your model. > > -- > Marcos Caceres
Received on Monday, 14 January 2013 20:25:14 UTC