[Standards] shared XML editing update

Peter Saint-Andre stpeter at stpeter.im
Mon Feb 4 23:25:07 UTC 2008

Boyd Fletcher wrote:
> On 2/4/08 6:09 PM, "Peter Saint-Andre" <stpeter at stpeter.im> wrote:
>> Boyd Fletcher wrote:
>>> Access control is a big issue for collaborative whiteboarding as you have
>>> read/write users, read only users, and users (like presenters) that can
>>> control the presentation (i.e. page flips and who is writing/changing the
>>> whiteboard). 
>> That should be handled by the floor control system (MUC or something
>> else), not the editing technology (SXE or something else).
> I disagree. The whiteboarding component needs to understand what is going so
> that it can maintain an accurate state of the whiteboard.


>>> Also, whiteboards are more than just a linear document of unrelated parts.
>>> It has a very well defined structure and if a client sends out an
>>> inappropriate/incorrect set of XML it can corrupt the entire whiteboard. SXE
>>> can¹t prevent that but a pure server based implementation can.
>> SXE in server mode can, but the spec does not yet define that in detail
>> by showing server-mode examples. That can be fixed by beefing up the spec.
>> Furthermore, SXE is now only the core editing protocol. Whiteboarding,
>> collaborative document editing, and other technologies can be layered on
>> top of SXE (the spec examples now show XHTML, not whiteboarding).
> that is a far more complicated approach so now you have to build a server
> side SXE component then build a whiteboarding component that jacks into the
> SXE component so that it properly handles the multitude of exceptions for
> the whiteboard? sounds messy to me.

What are all the exceptions for whiteboarding that don't apply to, say,
XHTML editing?

>>> we (and many large corporations) have operational experience and
>>> requirements for large numbers (>100) of people collaborating actively in a
>>> whiteboard session so not being able to handle that is a big deal.
>> See above. SXE should be workable in server mode. It's a question of
>> spec'ing that out more fully and building a working implementation that
>> extends the client-to-client mode with a server in the middle that
>> checks and (if necessary) corrects the XML.
>>> so basically, I don¹t see many advantages for the SXE approach for
>>> whiteboarding but I can see lots of disadvantages. I would seem to use that
>>> a server based implementation for whiteboarding is far more flexible for the
>>> types of sessions users really use whiteboarding for.
>> That is unproven.

I'm saying that it is possible to build a server-based implementation of
SXE. I mis-spoke by conflating your criticisms of SXE with your
criticisms of non-server-mode SXE.

> no its not. we have been using a server based component for a year and half
> with very good success.
> the SXE hasn't been proven to work for complex whiteboarding. our approach
> has. btw, where is a current working implementation of SXE or SXE based
> whiteboarding?

There is no implementation yet of server-mode SXE as far as I know. All
implementations are client-mode so far.

Joonas is going to update the SXE proposal to more fully document the
server-mode approach. Once that is done we can continue this
conversation. Until then all this discussion is pretty much moot.


Peter Saint-Andre

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 7338 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://mail.jabber.org/pipermail/standards/attachments/20080204/46c9a701/attachment.bin>

More information about the Standards mailing list