- From: Anne van Kesteren <annevk@opera.com>
- Date: Sat, 25 Jun 2011 19:07:40 +0200
- To: "Ian Hickson" <ian@hixie.ch>
- Cc: "Jonas Sicking" <jonas@sicking.cc>, "www-dom@w3.org" <www-dom@w3.org>
On Sat, 25 Jun 2011 19:00:04 +0200, Ian Hickson <ian@hixie.ch> wrote: > On Sat, 25 Jun 2011, Anne van Kesteren wrote: >> The way it is defined now is that events have a set of concepts >> associated with them and the object members return those. E.g. .bubbles >> "internally" is the bubble flag. It does not make sense if the event >> dictionary would refer directly to the properties. I suppose we could >> revamp the way events are defined though, but I am not sure how. > > Oh well "bubbles" and the other core attributes can be defined that way, > sure, they're more complicated. But for other events, there's not really > any magic in the attributes, right? They're read/write generally and > simply return the last thing they were set to. For those, no need for any > special flags. Well, they are as simple as bubbles I would say. E.g. ProgressEvent.lengthComputable is pretty much identical (currently its internal name is "maximum known flag"). How would you define http://dev.w3.org/2006/webapi/progress/#interface-progressevent in a more simple manner? Defining the mapping for the dictionary is rather trivial, but I agree that if we can make it more trivial or unnecessary even, that would be nice. -- Anne van Kesteren http://annevankesteren.nl/
Received on Saturday, 25 June 2011 17:08:13 UTC