W3C home > Mailing lists > Public > public-hydra@w3.org > December 2014

Re: Local identifiers in IriTemplates

From: Dietrich Schulten <ds@escalon.de>
Date: Mon, 22 Dec 2014 14:55:26 +0100
Message-ID: <549822CE.5050007@escalon.de>
To: public-hydra@w3.org
Hash: SHA1


For the issue of local identifiers, I am content so far.

Two comments below.

Am 21.12.2014 um 23:11 schrieb Markus Lanthaler:
> On 19 Dez 2014 at 11:37, Dietrich Schulten wrote:
>> Am 16.12.2014 um 23:08 schrieb Markus Lanthaler:
>>> On 15 Dez 2014 at 10:31, Dietrich Schulten wrote:
>> Right now what we have is the Traverson approach: you tell the
>> client to follow a rel path to achieve a goal. That is already a
>> huge change, compared with the traditional way. But my mid-term
>> goal is a client which I can tell to achieve a goal based on some
>> predefined knowledge, and it finds its way through a
>> hypermedia-enabled api.
> Remind me of our discussion in Paris :-)

Yes, I have been pondering that idea since "Follow your nose in
hypermedia" [1].

>> Part of that predefined knowledge would be "necessary
>> identifiers".
> And how would those identifiers be stored in the client (before
> interacting with the service the first time)?

Certainly at least per resource type and site, restricted to a
principal. (Identifier for schema:Event from http://example.com/api
owned by ds@escalon.de has value 4711 -> customer service
agent@example.com uses client to cancel event 4711 on behalf of the
event owner ds@escalon.de).

I wouldn't want a heuristic client (i.e. a client which only is likely
to do what it should) to change my resources either :-)

Best regards

[1] https://groups.google.com/forum/#!topic/hypermedia-web/HHXFoBh95ZQ	

- -- 
Dietrich Schulten
Escalon System-Entwicklung
Bubenhalde 10
74199 Untergruppenbach
Version: GnuPG v2.0.22 (MingW32)

Received on Monday, 22 December 2014 13:55:58 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:29:44 UTC