Hi,
Here is what we did :
http://clement.plop-plop.net/comet/
As you will see, it's quite simple, but has quite well worked.
Regards,
Clement
On 28.08.2009, at 09:26, Sten Roger Sandvik wrote:
Looked quickly at cometd and especially cometd-java. It looks to me
that
this is trivial to include in the http service implementation. Will
look
into this.
On Fri, Aug 28, 2009 at 9:03 AM, Sten Roger Sandvik <s...@x3m.com>
wrote:
The new http service is not testet alot. It's only been used in our
own
projects for now, but I will need to create more unit tests and some
integration tests. As for comet support - I have tought of it, but
have not
come around to do it. I will gladly look at the "current" comet
support to
see if it's a trivial ting to include.
Yes, it would be great to be included as a committer. I really like
the
Felix project and are really committed to create the best http
service
implementation out there :-)
BR,
Sten Roger
On Fri, Aug 28, 2009 at 8:52 AM, Rob Walker <r...@ascert.com> wrote:
Good point - we also have a "home grown" cometd approach which we
use for
server push to our GWT application, so something built into the
http server
would definitely be of interest
- R
Clement Escoffier wrote:
Hi,
Just my 2 cents.
any plan to support Cometd ?
We slightly change the current HTTP Service to support Cometd. I
don't
see any issue to do the same on Sten's version.
(of course, I can send what we quickly did).
As a reminder, Cometd is an HTTP based MOM, allowing (after a
handshake)
a server to notify browser. In Ajax-based interaction, the client
(periodically) query the server. With Comet, the server notifies
the
clients.
As you can imagine, this is definitely important for dynamic web
application.
Regards,
Clement
On 28.08.2009, at 08:40, Rob Walker wrote:
What do we need to do make this happen guys?
Sten - sounds like you've done a great job, kudos.
Felix/Marcel - sounds like you guys are happy with the approach
and code
From our side, we can certainly run some "real world"
compatibility
tests - which isn't to say in fact that Sten's new service would
be wrong
and the current one right, just whether thenew version breaks
anything
present in the exisiting service. If it does, I'm sure it'll
either be
fixable, or something that's actually not correct in the current
service -
so not a major issue. It'll be useful though to be able to
advise other
Felix guys of anything that might differ and need application
changes. I'm
happy to make time to look at this next week.
After that - do we just call a vote? I'm guessing Sten, we also
need to
propose you as a committter for maintenance?
Great progress though - the current Http service has served us
pretty
well, but it's always been on the list to have cleaner and
fuller solution
-- Rob
--
Ascert - Taking systems to the Edge
r...@ascert.com
+44 (0)20 7488 3470
www.ascert.com
