Re: WGLC of draft-ietf-xmpp-websocket-02

On 6/6/14, 1:47 PM, Ben Campbell wrote:
> Sorry for the delay. I just returned from a very disconnected vacation.
>
> On May 27, 2014, at 1:31 PM, Peter Saint-Andre <[email protected]> wrote:
>
>> Connection managers are trusted server components, so the administrators of the service are aware that there is a CM in the mix, but the server need not be (all it knows is that sessions have been created and that some trusted entity did so).
>>
>>> Is it assumed to
>>> implement this draft?
>>
>> The server doesn't really need to implement the websocket aspect of things, since that's handled by the CM.
>>
>>> Are we requiring special behavior of the
>>> server, without the server knowing it needs to do it?
>>
>> Not as far as I can see.
>
> Is there any reason to mention connection managers at all? Unless I missed something, the draft only mentions them parenthetically, but it does so in a normative statement. It's starting to sound like an implementation detail that doesn't need anything normative, if any mention at all.

It is in fact an implementation detail.

Peter