Dear All,
I'm glad to say I've had some good feedback from the last draft, which I've incorporated into a new document. Its only fair I circulate this first as there are one or two functional changes, although its on new functionality. The rest are mainly corrections:
* Have corrected references to FQDN and RadioVIS host and port to make them more distinct.
* Clarified restriction on ServiceIdentifier field to be maximum characters
* Clarified distinction between headers with a RadioVIS frame and HTTP headers
* Reversed time-direction of frames to be from older->newer. This made more sense with the restriction on number of frames and total response size.
* Corrected 4.5.1 where the trigger-time header was stated as being mandatory. This should have been optional. Whoops.
* Loosened up requirements on Stomp version to allow support for versions of Stomp later than 1.0, given the version negotiation in version 1.1.
* Clarified that a Stomp frame in its entirety be UTF-8, rather than just the message body
You can find this version of the specification here:
Since I've changed the meaning in some places I feel it only fair I post it here for your review. Please do check this over - I suspect many of you are getting draft spec fatigue by now, but I think we're nearly there...
I'm gonna say any feedback should be received by 09:00 GMT Friday 13th April. If you need more time then please say.
Ben