Right now I'm just getting text down that describes the basic points
without worrying too much about 100% correct spec text.  Take the actual
wording that's there with a grain of salt and wording suggestions are
greatly appreciated.

- James

Bjoern Hoehrmann wrote:
> * James M Snell wrote:
>> PaceSecurityConsiderations has been updated for Draft-09.  The various
>> MUSTs have been removed.
>>
>>  http://www.intertwingly.net/wiki/pie/PaceSecurityConsiderations
> 
> I am not sure you make proper use of upper- and lower-case RFC 2119
> keywords, e.g. in the second paragraph you have "strongly recommended",
> why is that not RECOMMENDED? The "MUST" in that paragraph seems in-
> correct, it's a statement of fact, not a conformance requirement of
> the Atom protocol specification.
> 
> In the next paragraph wording like "strongly encouraged" is rather
> misleading; if you don't want it to be RECOMMENDED, this should use
> something like "recommended but not required".
> 
> In 14.2 " multiple pipelined requests on multiple connections" seems
> a HTTP security concern, not a Atom protocol one (you don't restate
> all HTTP security considerations, so why this one).
> 
> You lack some references, e.g. to explain what "XML Digital Signature"
> is (assuming this is not provided in the current protocol draft).
> 
> Has this been reviewed against RFC 3552 and RFC 2828?

Reply via email to