Re: Let's get started

Hi.  I'm Kevin.

I've spent a good portion of my nerdy career tackling complexity in
enterprise software during the day and building infrastructure software by
night.  I've experimented with scaling both over a distributed architecture
at various organizations.

I've had roles in software development, enterprise architecture, process
improvement, and innovation.

I write open source software.  I like standards.  I used to have a giant
heart around XMPP, but she stopped calling me.  I live near HTTP these days.

I authored a hypermedia type named Siren, or application/vnd.siren+json for
you MIME lovers out there.  It's still in development.  Here it is:
https://github.com/kevinswiber/siren

I like studying distributed database systems and have started making steps
toward building my own.  I have experience in Domain Driven Design and
Behavior Driven Development.  I'm a programming language nerd.  I
understand the needs of the enterprise.  I have developer empathy.

Hydra is interesting.  In Siren, I've begun using class values as sort of
"type mixins."  An entity of a certain class contains attributes associated
with that class definition.  This is similar to Hydra types, except without
the Highlander principle applied.  In a Siren entity, there can be more
than one.

Siren has actions.  Actions are named, like forms.  I think these are
powerful.  I think Hydra maps "form inputs" to types in the vocabulary.
 Siren doesn't do this.  Reasons why or why not to do this would make a
good discussion.

I feel we are on the brink of something truly incredible in the hypermedia
space.  I believe in collaboration over competition.  It's too early to
claim #WINNING.

In general, I would like to see discussion here put in terms of the
_benefits_ REST provides, as opposed to subtleties in the definition of
"What is RESTful?"  Ain't nobody got time for that!

Cheers,

Kevin



On Thu, Jun 27, 2013 at 12:20 PM, Markus Lanthaler <markus.lanthaler@gmx.net
> wrote:

> Hi folks,
>
> The group has now 15 participants and probably a few more lurking on the
> mailing list, so let's get started. I think the best way to begin is to
> briefly introduce ourselves and to discuss our goals in order to flesh out
> a
> small roadmap.
>
> Some of you probably already know me through various related mailing lists
> or my work on JSON-LD. I have been a Web developer for more than a decade
> now. In the beginning I built small, static web sites but decided very
> early
> on that something more dynamic is needed. So I developed my first CMS about
> the time of Drupal's first release - who hasn't implemented a CMS? :-)
> Since
> then I've programmed almost everything from microcontrollers on Smart Cards
> in assembler up to large-scale distributed systems.
>
> More recently, not least due to my brilliant idea to do a PhD, I focused
> more and more on RESTful services and Linked Data. The result of this
> research is Hydra and JSON-LD (of which I'm one of the core designers and
> co-editor of the specifications). My (truly humble) hope is that they will
> form the base for a thriving ecosystem for machine-to-machine communication
> just as HTML does for the human Web.
>
> In the long term, I envision Hydra as set of modular vocabularies - that's
> the reason why I called the current spec "Hydra Core Vocabulary". There are
> a couple of things I would like to explore, for example:
>
> - support for binary data (this may be as simple as creating specialized
> classes)
>
> - authentication/authorization (including things like quotas and rate
> limits)
>
> - data validation (declarative description of the criteria for valid data,
> can be used to generate client-side and server-side validation code)
>
> - "single-click actions", i.e., operations which do not require any use
> input, such as a "like". Typically, I consider such things an anti-pattern
> in terms of a RESTful architecture but sometimes they are really handy
>
> Another thing I hope we can achieve is to simplify the core vocabulary even
> further.
>
>
> What are your goals? What are your pain points with current approaches?
> What
> do you think is still missing in the current core vocabulary?
>
>
> Cheers,
> Markus
>
>
> --
> Markus Lanthaler
> @markuslanthaler
>
>
>


-- 
Kevin Swiber
Projects: https://github.com/kevinswiber
Twitter: @kevinswiber