Re: [drinks] WG poll on transport protocol

[Theo Zourzouvillys <theo at crazygreek.co.uk>]

On Fri, Jun 26, 2009 at 2:29 PM, Alexander
Mayrhofer<alexander.mayrhofer at nic.at> wrote:

> The central counter argument is that SOAP is the most universally
> deployed data exchange mechanism out there

are you sure? I don't think i've needed to touch a SOAP API since
perhaps 2003.  Every single API around these days seems ot be RESTful
...

  http://www.programmableweb.com/apis/directory

But then, i've been working in an industry that doesn't move at
0.00034km/h below snails pace since then.

SOAP appears to be popular in corporate cultures as it is all
"point-clickey".  WSDL allows stub generation, integration is all very
pretty and useful if the developer has no clue what they're doing, but
as soon as you grind down to the protocol, it's seriously bloated and
relies on large libraries to do all the voodoo for you.  WADL now can
provide similar functionallity to WSDL, so i fail to see any excuses
for using SOAP anymore.

on a more serious practical note, people commonly stumble on using
REST over SOAP because they don't see the equvilent of WS-Transactions
and WS-Notifications in REST.  That doesn't mean you can't achieve the
same results, just it's not a "primitive" of the protocol like it is
in SOAP.

> 1. How many folks have SOAP implementations in their OSS-BSS systems for
> the
> kinds of protocols DRINKS is looking out?

probably a large number of the old-school carriers and telcos.

> 2. How many folks have a REST-like implementations in their OSS-BSS
> systems
> for the kinds of protocols DRINKS is looking out?

some do, for sure (we do), and i know of at least one multi-national
voice carrier than does.  The rest of the interwebs seem to prefer it,
too.

> 3. In general what data model or protocol properties do your
> organizations
> ultimately lean towards - SOAP vs. REST? Other?

REST

> 4. What are the specific applications that SSP are looking for DRINKS to
> enable .. LNP?..MMS/SMS interoperability, SIP interconnection etc ??

SIP inteconnection.

> Questions to reach WG concensus on:
>
> 3. Do you agree that the design team should define the architecture /
> data model independently from the definition of a potential
> transport protocol? (as far as possible)

obviously, to a degree.

While sending SOAP messages over SMTP seems "cool", and i'm sure it's
got a use, i frankly couldn't care less about it :-)

> 4. Given the goal to define the data model outside its transport, do you
> prefer the group to
>        A. Define a SOAP based transport protocol and if anyone is
> interested they can work on a RESTful version?
>        B. The WG try and do both?

putting my pragmatic hat on for a second, i'd prefer (A) over (B) if
the WG has no REST experience - It's harder to get SOAP wrong, as it's
already wrong -  otherwise, i'd prefer C: REST only.

 ~ Theo