Why we should link up our Web APIs

[Duncan Cragg]

I've put up a kinda sales pitch for the Object Network as part 2 of
the series.

It's here: http://duncan-cragg.org/blog/post/why-we-should-link-our-web-apis-object-network/

What do you think? Just me on my own, or about to spark a flood of
interest? Or somewhere in between?

Cheers!

Duncan

[Duncan Cragg]

Just published the next part: a kind of 'RFC' for the Object Network,
covering basic use of HTTP and JSON:

http://duncan-cragg.org/blog/post/basics-object-network/

On Jan 19, 9:23 pm, Duncan Cragg <duncan.cr...@gmail.com> wrote:
> I've put up a kinda sales pitch for the Object Network as part 2 of
> the series.
>

> It's here:http://duncan-cragg.org/blog/post/why-we-should-link-our-web-apis-obj...

[Jim Barritt]

I was going to ask about that Would you consider a full blown rfc or
is it too much like beauracracy?

Ar!

J

> --
> ________________________________________
> http://groups.google.com/group/forest-object-web

[Jim Barritt]

I think the vision is powerful. The example worked well for me. I
think having some community participation in designing the Interaction
protocol will help it to be accepted.

Perhaps it would be possible to write object network adapters for a
set of services (like the ones you mention)? This could kick start the
network and give a group of developers practical experience working
with it.

We could start that pretty quick. Have you got an object network
GitHub organisation set up?

Cheers

Jim

> --
> ________________________________________
> http://groups.google.com/group/forest-object-web

[Duncan Cragg]

> I was going to ask about that Would you consider a full blown rfc or
> is it too much like beauracracy?
>

It's best to follow the IETF mantra of 'rough consensus and running
code', before attempting an RFC. I've seen IETF standards work take
years for what should be pretty simple stuff. If everyone just agrees
in advance - because their code works together - it all becomes much
easier. Much easier to win arguments when you can point to running
code that wouldn't run properly if stupid decisions were made. Much
easier to talk about issues in the code rather than in the abstract.

I'm sure that's what you were afraid of in the 'bureaucracy'
thing.. :-)

I do see a base media type coming out of this.

But I've seen others put their stall out on a one-man RFC and expect
people to jump on board. Na. Not here.

Cheers!

Duncan

[Duncan Cragg]

> I think having some community participation in designing the Interaction
> protocol will help it to be accepted.

Absolutely. As long as I get the final veto! :-) OK, ultimately, no-
one can stop a bunch of people going off and doing what they want, and
the power of the browser builders over the W3C is an example of that.
So replace 'some' with 'plenty of'.

>
> Perhaps it would be possible to write object network adapters for a
> set of services (like the ones you mention)? This could kick start the
> network and give a group of developers practical experience working

> with it. .. We could start that pretty quick.

Oh yes. I've been reading 'Weaving the Web' by Tim Berners-Lee!

I see all that Semantic Web Linked Data sitting there unloved, ready
to be brought to life in interlinked JSON!

There's a lot of data in a lot of APIs that can be freed up into the
Object Net. Anyone tried ql.io?

> Have you got an object network GitHub organisation set up?
>

Not yet. I've got a whole load of code that needs some lovin', though.
Javascript, Node.js, Java, Android...

What exactly did you have in mind there?

Cheers!

Duncan