Re: [schemeProtocols-49] New draft of proposed "URI Schemes and Web Protocols" Finding

On Monday 2005-11-21 22:04 -0500, noah_mendelsohn@us.ibm.com wrote:
> This note is to announce that I have prepared a significant 
> revision to the draft finding on "URI Schemes and Web Protocols" [3], and 

Two comments on
http://www.w3.org/2001/tag/doc/schemeProtocols-2005-11-21.html :


3.5 (Scenario: A new scheme to launch a media player) is tied to
deficiencies in desktop environments.  Major desktop environments today
(Windows, Mac OS X, GNOME, KDE) can be relied on to have two mechanisms
for dispatching information resources to applications: dispatching the
representation based on the media type, or dispatching the URI based on
the scheme.  That the environments don't generally give a way to
dispatch the URI on the basis of the media type causes people to invent
new URI schemes (e.g., webcal:, itms: [1]) because it is the simplest
way to dispatch a URI (rather than or in addition to a representation)
to an external application.  (RealPlayer's solution to this problem
predated reliable scheme dispatch, though, so it uses a separate file
containing a URI.)

Thus another set of guidelines may be needed related to this example:
those for platforms on which applications that use resources from the
Web are written.  Their mechanisms for dispatching information resources
to applications based on media types should allow dispatching the URI,
the representation, or both.


3.3 (Scenario: Accessing https resources using a peer-to-peer protocol)
and 4.1.3 R3 (URI scheme sets expectations for integrity of access) both
could be interpreted to imply that https is used only for
authentication.  It should be clear that it's used for encryption as
well.

-David

[1] http://lists.w3.org/Archives/Public/www-tag/2003Apr/0151

-- 
L. David Baron                                <URL: http://dbaron.org/ >
           Technical Lead, Layout & CSS, Mozilla Corporation