Dear All,
I'm following up the email containing the latest draft with a view on the points I think need adding to the RadioEPG specification, based on comments I have gathered.
This is by no means exhaustive, and I've probably missed out on some, having really only come to this in an official capacity recently. If you have any outstanding concerns with what is in the specification then please do let me know as soon as possible.
Here are the points I shall be looking at:
* Guidance for Gzip compression of documents
* A method for requesting and returning a more appropriate XSI file (think Content Negotiation for XSI)
* Are the notes for Service Following sufficient?
We held up ratification at the last GA partly because it was deemed that insufficient discussion had taken place, and that there a few remaining issues.
The first was a concern I had that indicating bearer offsets in the XSI would not be the best way for devices to synchronise across platforms. However, I deem this value to be an 'indicative' rather than absolute value and so can be useful for bearer selection and a rough figure of offset. Work is continuing separately to explore means of better synchronisation between bearers and RadioDNS applications.
Another point was raised to do with how bearers are represented in RadioEPG - I originally put on my semantic hat and used the strict URN syntax. Others favoured the URL syntax, which I resisted due to the fact that the bearer is not a URL as it does not point towards a locateable resource. I havent' had any significant followup from this discussion, but I'd like to know whether anyone feels this is a point worthy of further discussion.
I'll post additional messages regarding the three points above, but PLEASE do raise any issues you may have, or things that still dont feel right about this specification. I'm hopeful we can still get this to be ratified at the next RadioDNS GA, but will need some assistance to make this happen.
Kind regards,
Ben