[jdev] Facebook XMPP

Nick Vidal nick at inf.ufrgs.br
Thu May 15 15:28:06 CDT 2008


How can Facebook (and others) win by adopting XMPP to its full potential?

If we can answer this question and write an open letter to Facebook, Google,
Yahoo, Microsoft, Twitter, etc, successfully making them realize that this
is the way to go, inviting them to have access to these valuable resources
created by the XSF, then we all win. Are lock-in strategies still benefitial
for them in this new scenario? I don't believe so, and I believe we can
convince them of that. So how about writting an open letter to these
influential companies? Who thinks this is a good idea?

On Thu, May 15, 2008 at 4:44 PM, Scott Lewis <slewis at composent.com> wrote:

> Hi Folks,
>
> (Lurker materializes)
>
> One comment I would like to make about this discussion of whether or not to
> work on multiprotocol clients/i.e. whytransportsmatter.
> It's not realistic IMHO to expect that the whole world will transfer to
> open protocols/XMPP overnight...as much as some of us would like to see this
> happen.  Rather I think the key to making this happen is make such
> transitions as easy as possible...by:
>
> 1) Having lots of clients (whether single protocol or multi-protocol) so
> that UI innovation can occur and create new user value
> 2) Having lots of good clients
> 3) Having open clients (open protocol at least...and preferably open source
> implementations)
>
> As important as it is, I think it's still very hard to convince users that
> they should choose interoperability over UI features.  So for
> interoperability to matter to users, open clients have to be as good,
> numerous, and innovative as well as support interoperability.  Further,
> multiprotocol clients can expose the value of interoperability to users
> while still giving them what they want:   easy/familiar connectivity to
> others.
>
> In order to help 1, 2, and 3 along, I/we have taken the approach of
> creating protocol independent communications APIs as part of the ECF
> project:  http://www.eclipse.org/ecf.  It's our hope that by creating a
> protocol-independent, open and extensible 'presence' API (as but one
> example) it makes it possible for developers to create either single
> protocol or multi-protocol clients more easily/quickly/with higher quality,
> and without taking a least common denominator approach to features (because
> both the core and all ECF APIs are extensible at runtime via OSGi either in
> servers or client applications).
>
> Also, such an approach minimizes the effort in creating multiprotocol
> clients...not that it doesn't eliminate it, but it does reduce it to a more
> manageable level for client developers.
>
> Anyway...I'm happy about the Facebook announcement too :).
>
> Scott
>
>
>
> Sander Devrieze wrote:
>
>> 2008/5/15 Nick Vidal <nick at inf.ufrgs.br>:
>>
>>
>>> Sanders: you do support users who use AIM and MSN, since you *waste your
>>> time* making sure coccinella works with transports. And you do support
>>> users
>>> of Microsoft Windows, since you *wast your time* making sure coccinella
>>> works in Windows. And this is a good thing! Thank you! :)
>>>
>>>
>>
>> My reply is here as already said before:
>> http://coccinella.im/whytransportsmatter
>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.jabber.org/pipermail/jdev/attachments/20080515/711708fd/attachment-0002.htm>


More information about the JDev mailing list