Re: [Beepcore-java-users] PiggyBacked data, MessageListener and RequestHandler (simple fix)

[William J. M. <wm...@es...>]

Vladimir,

I had hoped Huston would respond.  I am not strong at all
on the Java implementation, so I was staying to the side.

If you don't mind, please let us know how this works after you have
shaken it out a bit, and then perhaps we can see about getting it
checked in, and hopefully Huston will re-surface and can comment on
how the API would be affected.

I suspect your change is probably right.

Regards,

-bill mills


On Wed, Feb 11, 2004 at 10:32:45AM -0500, Wladimir Araujo wrote:
> 	Hi,
>=20
> 	Since I did not hear from anybody in this respect, I'm adding this simpl=
e fix. The piggybacked stuff remains there and the TCPSession is recreated =
afterwards. This only works if StartChannelListener.startChannel() calls ch=
annel.setStartData() to send back a piggybacked reply, otherwise the sessio=
n will die if the new channel does not reply to the piggybacked message. Th=
e patch is below.
>=20
> 		Cheers,
>=20
> --- SessionImpl.java.new        Wed Feb 11 10:21:28 2004
> +++ SessionImpl.java.old        Wed Feb 11 10:24:09 2004
> @@ -1321,7 +1321,7 @@
> =20
>                  fireChannelStarted(ch);
> =20
> -                if (ch.getState() !=3D ChannelImpl.STATE_TUNING) {
> +                if (p.data =3D=3D null && ch.getState() !=3D ChannelImpl=
.STATE_TUNING) {
>                      this.enableIO();
>                  }
>              }
>=20
> Wladimir de Lara Araujo Filho
> Juniper Networks, Inc.
>=20
>=20
>=20
> -------------------------------------------------------
> SF.Net is sponsored by: Speed Start Your Linux Apps Now.
> Build and deploy apps & Web services for Linux with
> a free DVD software kit from IBM. Click Now!
> http://ads.osdn.com/?ad_id=1356&alloc_id438&op=CCk
> _______________________________________________
> Beepcore-java-users mailing list
> Bee...@li...
> https://lists.sourceforge.net/lists/listinfo/beepcore-java-users