RE: [TF-DI] Thing API proposal (was RE: [TF-DI] Agenda and webex details - 24 Sept 2015 at 15:00 CEST)

Hi Dave,

We are well aware of NSD API activity as we provided a WebKit implementation [1], and indeed some security issues could not be overcome with this model. Yet, Presentation API offers an alternative model that does not suffer from the same issues and could be extended to provide a broader discovery/communication API, as illustrated by Louay’s recent proposal but also by different presentations at TPAC last year [2].

I agree that clarifying use cases and getting support from browser vendors would be necessary to initiate new work on this topic, but based on related presentations, proposals and previous work, there seems to be an interest for such an activity. While different solutions may be envisioned, WoT already provides a compelling use case for a discovery/communication API as it would allow to set up a pipe between web apps and things, a means that WoT is currently missing and that could easily be extended (e.g. based on Thing Description and interaction models).

Best regards,

Romain.

[1] https://github.com/canonresearchfrance 
[2] https://www.w3.org/wiki/TPAC2014/SessionIdeas#startSession.28.22WoT_devices.22.29


> -----Original Message-----
> From: Dave Raggett [mailto:dsr@w3.org]
> Sent: mardi 13 octobre 2015 19:13
> To: BELLESSORT Romain
> Cc: Bassbouss, Louay; fd@w3.org; Hund, Johannes; public-wot-ig@w3.org;
> FABLET Youenn
> Subject: Re: [TF-DI] Thing API proposal (was RE: [TF-DI] Agenda and webex
> details - 24 Sept 2015 at 15:00 CEST)
> 
> 
> On 13 Oct 2015, at 17:49, BELLESSORT Romain
> <Romain.Bellessort@crf.canon.fr> wrote:
> 
> As a side note, I believe that a generic discovery/communication API would
> also be very useful for WoT. Indeed, in addition to providing a means which is
> currently lacking (as would Thing API), it would also allow developers to keep
> experimenting and adding custom features if needed thanks to its
> extensibility.
> 
> The W3C Device APIs WG has looked at a network service discovery API, but
> it failed to get sufficient support to progress further. Instead, we are seeing
> progress on discovery that is tied into APIs for particular services, e.g. the
> second screen API, and the work on NFC and Bluetooth (the latter two
> currently in Community Groups).
> 
> One of the issues is how to address privacy, another is the need to adhere to
> the Web’s same origin security policy, although CORS provides controlled
> way to relax this.  The privacy challenge is considered to be easier to address
> for specific services.  If we are to initiate new work on a generic
> discovery/communication API for the Open Web Platform it will be important
> to clarify the use cases, and to get support from browser vendors.
> 
> Best regards,
> —
>    Dave Raggett <dsr@w3.org>
> 
>

Received on Thursday, 15 October 2015 16:31:51 UTC