On Thu, Mar 3, 2011 at 5:07 PM, Peter Saint-Andre <[email protected]> wrote:
>> I have only one comment at the moment.
>>
>> I would not like to drop support for group-chat.

Group chat is doable as a private extension to any standard I will
introduce.  I might even work with specific parties who might decide
to introduce groupchat.     People here who received my "Supplement"
Word document showed that there were 2 or 3 pages of groupchat issues
that needed considering (I didn't mention those in the submitted
standard) -- email me privately for a copy.   I think at this stage,
it's pretty plainly clear that groupchat complicates the standard
sufficiently enough to downgrade it to a single-line bullet into an
"Future Experimentation" section.   It was never fully documented in
the submitted standard, and I warned about its experimentalness.

> (I would sometimes find chat state notifications useful, e.g. during a
> meeting of the XMPP Council when I really care whether one of the
> Council members is about to post a message to the room. Right now I
> can't think of a time when I would want real-time text in a chatroom,
> but I can envision that deaf people might think differently.)

If 10 people started typing at the same time, it will be quite
distracting.  But it does automatically discipline people to type only
one or two people at a time (if they were doing construction
discussion).  On the other hand, I can envision a "lecturer" /
"teacher" / "leader" being allowed to send real time text, with
everyone not sending real time text.   That would be useful in
disabled classroom situations, where anyone can pre-compose questions
and only send them if necessary.

That said, I do intentionally design my protocol to be compatible with
a theoretical / future / private groupchat extension.  I never really
defined how groupchat works, and when fully explaining the scenarios
(moving content from the Supplement document to the spec), it
increases the spec size to larger than the spec that I originally
submitted.

One of my two software packages utilizes a >2 realtime text groupchat,
so it will definitely use a private groupchat extension.  Right now,
it's running on a proprietary protocol but I'd love to convert it to
XMPP Real Time Text with a groupchat extension (even a private
groupchat extension).   This software I wrote, is a project that has
existed from long before I started writing the XMPP Real Time Text
spec.

If there's big demand for interop, we can consider a standardized
public groupchat extension later.
The point is that my standard doesn't preclude ability to do groupchat
extension.

Reply via email to