- From: Giuseppe Pascale <giuseppep@opera.com>
- Date: Thu, 24 Nov 2011 18:09:14 +0100
- To: "Paul Kinlan" <paulkinlan@google.com>, "Jean-Claude Dufourd" <jean-claude.dufourd@telecom-paristech.fr>
- Cc: public-web-intents@w3.org
I think we are using the term "discovery" with different meanings, hence the confusion. Web Intents allow an *application* to discover a *provider* of a certain service that was previously registered by the UA in some way(e.g. when the user navigated to a page that declared to be able to provide such service) HN discovery protocols allow the UA to discover devices that provide services without users and/or applications involved. What we should do is connect these 2 things so that intents can be used to expose to an application a service that the UA found on the home network. Actually the same can be done on the web, if you happen to have a Search engine that can look around for <intent> tags ;) Of course in a scenario where they do not explicitly register an intent provider (like going on the page that provides a service) you need to workout another way to preserve privacy and security. This needs to be handled by the UA. /g On Thu, 24 Nov 2011 17:02:57 +0100, Jean-Claude Dufourd <jean-claude.dufourd@telecom-paristech.fr> wrote: > This is exactly why I said that we may have a different definition of > discovery. > You seem to be interested mainly in discovering services that are not > yet known but are directly accessible in the environment of the current > device (read: that do not need an (IP) address to be talked to): > - peripherals connected to the current device; > - optional software installed on the device... > The Home Network TF is only interested in discovering services elsewhere > on the network, on devices unknown to the discoverer. > > And even your version of discovery is outside the scope of Web Intents > if I am not mistaken. > Best regards > JC > > On 24/11/11 16:01 , Paul Kinlan wrote: >> We don't have any addresses in the discovery process, the >> webintent.org/share urls are a simple string that is used to allow the >> UA to know the task the user want't to achieve. It could be just a >> plain string "read-temperature" >> >> Our current documents describe the process of registration that can be >> detected by the UA, via the parsing of the intent tag, but we haven't >> said that they only way that a UA can know about a service is via only >> the intent tag. >> >> As Greg said earlier, we will be able to list the intents a service >> supports via the chrome extension/app mechanism - this is not done via >> the intent tag detection and is still conformant with the spec we had >> defined. >> >> P >> >> On Thu, Nov 24, 2011 at 1:16 PM, Jean-Claude Dufourd >> <jean-claude.dufourd@telecom-paristech.fr> wrote: >>> On 23/11/11 20:01 , timeless wrote: >>>>> All my confusion (shared by others I believe) came from trying to >>>>> understand >>>>> how Web Intents provided discovery, which was hopeless :) >>>> Sorry, it does provide *a* mechanism for Discovery of /some/ services, >>>> but the point is that it wasn't going to limit that to the only one, >>>> and that other ways to be Discovered could be provided. >>> JCD: We may have a different definition of discovery. >>> I have found nothing in the current spec that allows a web app to >>> discover a >>> service without having an address beforehand (whether the address is >>> that of >>> the service or that of a directory). Please correct me if you think I >>> am >>> wrong. >>> Best regards >>> JC >>> >>> -- >>> JC Dufourd >>> Directeur d'Etudes/Professor >>> Groupe Multimedia/Multimedia Group >>> Traitement du Signal et Images/Signal and Image Processing >>> Telecom ParisTech, 37-39 rue Dareau, 75014 Paris, France >>> Tel: +33145817733 - Mob: +33677843843 - Fax: +33145817144 >>> >>> >>> >> >> > > -- Giuseppe Pascale TV & Connected Devices Opera Software
Received on Thursday, 24 November 2011 17:09:58 UTC