Hi everyone,

Lots of great comments. Thanks. I will take them into account in the next flavor of the example I had written. The example given was indeed meant to be pseudo-code.

Fundamental points from our conversation so far: (a) the JSON request/response should be a reflection of the XML version. One can translate in either direction without any loss (b) Use of constants vs. use of assumed (default) values: I still think assuming values is nicer. It makes creating requests by hand so much easier. It reduces the size of a request as well and increases readability. (c) short attribute Ids vs. longer (fully qualified ones): Danny you say that shorter names would lead to possible clashes. But I have never run into that issue w/ customers who tend to live in a closed world where all attributes are clearly defined.

My example, in fact, should have skipped the datatype and the issuer (as Ray pointed out later).

I will get back to you w/ additional work in the coming weeks. Unfortunately, due to a prior commitment, I won't be able to attend today's call.

Cheers, David.

On Thu, May 10, 2012 at 3:15 PM, <remo...@emc.com> wrote: