Re: [apps-discuss] Fwd: New Version Notification for draft-shelby-exi-registration-01.txt

[Zach Shelby <[e-mail]>]

On Apr 12, 2012, at 11:57 PM, Ned Freed wrote:

>> So it seems like +exi *could* be used if the representation stuck to
>> only using the generic dictionary and eschewed any type-specific
>> out-of-band dictionaries.
> 
> Agreed. That would make the definition of +exi acceptable. Whether or not
> that's an acceptable restriction for users of EXI is, of course, another
> question entirely.

Using EXI in its normal mode is already fine as a content-encoding. The problem we need to solve here is exactly the Schema-informed mode where an indication of out-of-band dictionary information is needed, and where the use of content-encoding is not really appropriate (as we all seem to agree).

Now I am as much on the fence as Ned it here about this. Seems we are in a sticky situation where serializations that require OOB dictionary information fall between RESTful cracks. Examples of these that I know of include:

-  ASN.1 PER 
- EXI Schema-informed mode
- Google Protocol Buffers 
- Surely more... 

These tend to be applied in M2M and other applications where data efficiency is extremely important and using OOB dictionary information is what gives us that efficiency. 

Zach

-- 
Zach Shelby, Chief Nerd, Sensinode Ltd.
http://www.sensinode.com
http://zachshelby.org  - My blog "On the Internet of Things"
http://6lowpan.net - My book "6LoWPAN: The Wireless Embedded Internet"
Mobile: [phone]