Re: [sca-bindings] Issue 2 proposal v14

[Eric Johnson <eric@tibco.com>]

The only thing that gets me is the passive voice in the last sentence. 
The passive voice in the rest of the paragraph may be hard to avoid.

Anish Karmarkar wrote:
> Attached.
>
> The changes I have made from v13 are:
> 1) s/200712/200903/g to sync up with the assembly PR NS
> 2) s/\/200812//g -- this is for the relationship URI. I don't think we
> need a date component in here. This would avoid any confusion wrt the
> NS URIs. In fact the recent trend is to not include the date in the NS
> URIs either. See WS-Policy NS URI. I know OASIS has a NS guideline
> which has YYYYMM in it, but this isn't a NS URI, so the policy doesn't
> apply.
> 3) I changed the wordings for the listener case (based on Mike's
> suggestion). It reads as follows:
>
> "When a service that offers a bidirectional interface is invoked,
> depending on the semantics and/or implementation of the service, it is
> possible that the service might invoke the callback interface before
> the forward operation ends. In such cases, it is necessary for the
> binding on the reference-side to be listening for callback request(s)
> from the service, before the forward operation request is sent on the
> wire to the service, and continue listening as long as callback
> requests are expected. It is possible that before the response to the
> forward request is sent a response to one or more callback requests
> are required by the service."

How about "As one possible example, a service could send one or more
callback requests before determining and sending the response the
original forward request."
>
> I'm hoping that we can close the issue on the call.
Me too, which makes me add that I do not want to hold up the proposal
either way, and equally makes me hesitant to even suggest this small
change.  I'm very sorry I didn't suggest it sooner.

-Eric.