W3C home > Mailing lists > Public > public-hydra@w3.org > April 2016

Re: Interesting discussion on Twitter

From: Ruben Verborgh <ruben.verborgh@ugent.be>
Date: Wed, 27 Apr 2016 10:25:38 +0200
Cc: Hydra <public-hydra@w3.org>, darrel.miller@gmail.com
Message-Id: <986D9503-F29E-4CE7-AB38-AED84A37E86C@ugent.be>
To: Markus Lanthaler <markus.lanthaler@gmx.net>
HI Markus,

Interesting discussion.

This part should definitely be discussed with Darrel:

> I am not convinced of the value of HYDRA. Its mechanisms to augment representations with links seem redundant to me and I don't see the value of yet another API description language that encourages use of HTTP in tightly coupled way.

@Darrel: I'm not sure whether you fully get Hydra then.
Hydra is a vocabulary that allows you to describe hypermedia controls.
It's like the <form>, <input>, and <label> functionality of HTML.
What's tightly coupled about that?

And given the Twitter conversation, what's your alternative to realize that functionality?
>> "The predicate is the link relation type Standard LRTs don't have URIs, so you would need to prefix with some base URL"
https://twitter.com/darrel_miller/status/724339624345276417
Well, Hydra provides such a base URL.
That not tightly coupled compared to everybody defining their own.

(There are other parts of Hydra that focus on API descriptions,
but that does not necessarily imply tight coupling.)

So I think that Hydra is being, at least partly, misrepresented in the interview.
And that's too bad, because from experience with many formats for APIs, I can tell that
Hydra is really doing something different since it works on in-band parts of the message.

Best,

Ruben
Received on Wednesday, 27 April 2016 08:26:10 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 27 April 2016 08:26:10 UTC