- From: Erik Wilde <dret@berkeley.edu>
- Date: Tue, 28 Jul 2015 17:03:23 -0700
- To: Annette Greiner <amgreiner@lbl.gov>, Andreas Kuckartz <a.kuckartz@ping.de>, Data on the Web Best Practices Working Group <public-dwbp-wg@w3.org>
- CC: Carlos Iglesias <carlos.iglesias.moro@gmail.com>, Newton Calegari <newton@nic.br>, Yaso <yaso@nic.br>
hello annette.
On 2015-07-26 23:53, Annette Greiner wrote:
> Does anybody have an example of a successful HATEOAS compliant API that does not use versioning?
> I'm honestly wondering, not trolling here. In order to argue for something as a best practice, it helps to be able to point to successful use.
well, i hope this doesn't sound too obvious, but HTML is a great
example. you can use old pages with new browsers and new pages with old
browsers. if HTML would have designed in a way where each "version" of a
page would have required different URIs, and consumers somehow
navigating between those, it's easy to imagine that the web's success
would have been adversely impacted.
in terms of non-HTML examples, there are numerous examples of spec-level
formats that have gone that way. two examples that come to mind ate Atom
and DITA. both have extension and openness models that allow users to
evolve the formats with additions (for example, podcasts are nothing but
an extension of Atom, but not "a new version"), allowing the ecosystem
of producers and consumers to evolve and grow without creating
versioning rifts (that tend to impact this kind of organic growth).
cheers,
dret.
--
erik wilde | mailto:dret@berkeley.edu - tel:+1-510-2061079 |
| UC Berkeley - School of Information (ISchool) |
| http://dret.net/netdret http://twitter.com/dret |
Received on Wednesday, 29 July 2015 00:03:54 UTC