W3C home > Mailing lists > Public > public-webapps@w3.org > July to September 2011

Re: [WebIDL] remove modules

From: Paddy Byers <paddy.byers@gmail.com>
Date: Fri, 12 Aug 2011 23:43:28 +0100
Message-ID: <CABEaZavPXew5xDC3YPniL-JSzAtBdtm0vqoLAhAxXO7-6wsNQA@mail.gmail.com>
To: Arthur Barstow <art.barstow@nokia.com>
Cc: public-webapps@w3.org, public-script-coord <public-script-coord@w3.org>
Hi,,

If modules are removed from the Web IDL spec, what running code e.g.
> browsers, web/widget runtimes, IDEs, test cases, etc. will no longer comply
> with the spec (looking for real breakages here)?
>
> If WAC needs that type of functionality, could they define their own IDL
> extension?
>

Of course WAC can define something. The question is whether or not there is
general utility in having modules as a way of logically grouping a set of
interface definitions.

The thing we do that will break is that a module is used to group the
interfaces that are associated with a particular feature. For example, when
a programmer requests a feature - say
http://www.w3.org/TR/geolocation-API/- then the interfaces that are
enabled are exactly those belonging to the
geolocation module. Without a module to group these interfaces, all we have
is that the feature URI and the interfaces are implicitly connected by being
defined in the same document.

Concretely, this is used in the WAC SDK. When a widget's config.xml includes
a feature element, then the interfaces and prototypes associated with that
module are instantiated for content assistance purposes. If we only had that
grouping implicit, or in prose in the spec, we would need to preprocess the
WebIDL ad add further metadata, in order to include each new API into the
SDK.

Thanks - Paddy
Received on Friday, 12 August 2011 22:43:59 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:49:47 GMT