W3C home > Mailing lists > Public > public-dwbp-wg@w3.org > December 2015

Re: dwbp-ISSUE-213 (BP_REST): Changes on BP about REST APIs [Best practices document(s)]

From: Yasodara <yaso@nic.br>
Date: Fri, 04 Dec 2015 10:09:32 -0200
Message-ID: <5661827C.9010509@nic.br>
To: Annette Greiner <amgreiner@lbl.gov>
CC: public-dwbp-wg@w3.org, Erik Wilde <erik.wilde@dret.net>
Hi Annete, Erik and everyone,

I made a lot of changes in the draft proposed by Annete, hope to make a pull request at github for the editors, but it would be nice if you could take a look.


https://docs.google.com/document/d/13q2CvZjMkjXrg7Pv7aTRuU5UL2J_-py09Vs_zJoaLso/edit#


yaso




On 11/23/2015 08:24 PM, Annette Greiner wrote:
> Hi Erik,
> I can live with avoiding using the word REST when it isn't used to
> describe an API that follows all the REST constraints, but there is a
> webby coding style that most web API developers follow that often makes
> use of hypermedia but doesn't rely on it for holding state. Developers
> using this style do expose resources with consistent IDs identified by
> URIs and manipulated via representations. What do you suggest we call
> that approach?
> -Annette
>
> On 11/23/15 12:26 AM, Erik Wilde wrote:
> > On 2015-11-20 20:51, Annette Greiner wrote:
> >> As a group, we have, I think wisely, avoided centering any of the best
> >> practices on any single technology, because we want the practices to be
> >> able to survive the forward progress of technology and popularity. We
> >> can mention specific technologies in the implementation sections,
> >> though.
> >
> > +1 for not recommending specific technologies, and +1 for maybe
> > listing some as examples if that's a pattern throughout the spec.
> >
> >> I also believe we need to consider pragmatic REST as well as HATEOAS.
> >
> > REST has a proper definition. "pragmatic REST" doesn't (or it's
> > something along the lines of "do whatever you like via HTTP, mostly
> > use it like FTP with a different name").
> >
> > HATEOAS *as a name* should be avoided, people actively try to avoid it
> > these days. call it hypermedia instead, and say that it's the essence
> > of webbyness.
> >
> >>> On Nov 16, 2015, at 11:52 AM, Pieter Colpaert
> >>> Instead I would suggest taking it a step further. When we combine
> >>> REST, HATEOAS, Linked Data and HTTP for an Open Data API thoroughly,
> >>> you'd come up with Hydra [1] (a community group [2] at W3C). Wouldn't
> >>> it be better to create a best practise that says you need to weave
> >>> your data documents together with Hydra?
> >
> > i concur with annette when it comes to not recommending specific
> > technologies. also, you can make much more lightweight recommendations
> > than this one, which means that people have to use the RDF metamodel.
> >
> > what might be a very lightweight and pragmatic best practice about
> > hypermedia would by to say that one of the essential aspects of
> > hypermedia are typed links:
> >
> > https://github.com/dret/hyperpedia/blob/master/concepts.md#link-relation-type
> >
> >
> > and that when it comes to typed links, RFC 5988 and the resulting link
> > type registry at
> > http://www.iana.org/assignments/link-relations/link-relations.xhtml
> > provide a simple foundation for anybody to build their vocabulary on.
> >
> > cheers,
> >
> > dret.
> >
>

-- 
-yaso

CeWeb - Centro de estudos sobre tecnologias Web
W3C Escritório Brasil
NIC.br - CGI.br
Phone: 55 11 5509-3537 (4025)
Received on Friday, 4 December 2015 12:10:09 UTC

This archive was generated by hypermail 2.3.1 : Friday, 4 December 2015 12:10:09 UTC