[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: CAP Capabilities



Doug Royer wrote:
> 
> > Date: Mon, 19 Jul 1999 15:11:05 -0700
> > From: sman@netscape.com (Steve Mansour)
> >
> > >
> > > > What do people think of this?  Any other options to consider?
> > >
> > > I don't see a conflict.
> > >
> > > Commands replies have 2 parts, the protocol reply, and optional
> > > application reply:
> > >
> > >         CUA: AUTHENTICATE .....
> > >         CS:  2.0 <or what ever is a valid reply>
> > >         CS:  <MIME OBJECT>
> > >         CS:  BEGIN: ...
> > >         CS:  ...
> > >         CS:  END: ...
> > >
> > > The 2.0 is the protocol reply, the MIME object that contains a
> > > BEGIN ... END is the optional application reply.
> > >
> > > Other commands have application data, that's the point of sending
> > > a command. Correct?
> >
> > Right.
> >
> > The concern was that in response to a transport command, we're returning both
> > transport data as well as application data.
> 
> Yes - the transport command is AUTHENTICATE or SENDDATA, what's the
> difference? Both return application data. I see this as consistency.
> 

When you issue a SENDDATA you send application level data with it. 
The application level data returned is in response to that.  When you
authenticate you are sending no application level data, so if you get
some back it is unsolicited at the application layer.  

Do we want to allow the receipt of unsolicited data?

Alex