Socket.IO-node 0.6.5
19 views
Skip to first unread message
Guillermo Rauch
Jan 9, 2011, 7:59:50 PM1/9/11
to sock...@googlegroups.com
Two important bugs that were introduced (regarding polling transports) and went unnoticed due to the automated tests not being able to simulate a real-world web browser connection have been resolved in 0.6.5.
Everyone is encouraged to upgrade, either by:
npm install socket.io
or checking out the latest `0.6.5` tag.
Sorry for the trouble, and special thanks to Sascha Gehlich for his help with the bug reproduction.
Allen Lee
Jan 10, 2011, 4:08:35 PM1/10/11
to sock...@googlegroups.com
Is this on github? I'm not seeing the 0.6.5 tag on either the master
or the develop branches...
or the develop branches...
Thanks for the code :-).
--
Allen Lee
Center for the Study of Institutional Diversity [http://csid.asu.edu]
Arizona State University | P.O. Box 872402 | Tempe, Arizona 85287-2402
Office: 480.727.0401 | Fax: 480.965.7671
Mike Pilsbury (pekim)
Jan 10, 2011, 2:31:54 PM1/10/11
to Socket.IO
I thought that I was going mad yesterday. I'd used socket.io
previously, with few problems, and I couldn't figure out what I was
doing wrong, and why the connect was closing after 7 or 8 seconds. I
simplified the code down to almost nothing, and still it wasn't
working.
As soon as I installed this update (actually 0.6.8) I was a lot
happier. Thank you.
Mike
previously, with few problems, and I couldn't figure out what I was
doing wrong, and why the connect was closing after 7 or 8 seconds. I
simplified the code down to almost nothing, and still it wasn't
working.
As soon as I installed this update (actually 0.6.8) I was a lot
happier. Thank you.
Mike
Allen Lee
Jan 10, 2011, 5:27:45 PM1/10/11
to sock...@googlegroups.com
Guillermo Rauch
Jan 10, 2011, 7:43:42 PM1/10/11
to sock...@googlegroups.com
Just pull the latest tag =]
Peter
Jan 10, 2011, 7:47:34 PM1/10/11
to Socket.IO
is 0.6.5 supposed to fix the disconnection issues on Android/iPhone?
the changelog was a bit vague and it seems like it improved it a bit
but I still see disconnections in the test case I sent you
On Jan 10, 4:43 pm, Guillermo Rauch <rau...@gmail.com> wrote:
> Just pull the latest tag =]
>
>
>
>
>
>
>
>
>
> On Mon, Jan 10, 2011 at 2:27 PM, Allen Lee <Allen....@asu.edu> wrote:
> > Nevermind, I just realized this is for socket.io-node and not the
> > client socket.io - doh!
>
the changelog was a bit vague and it seems like it improved it a bit
but I still see disconnections in the test case I sent you
On Jan 10, 4:43 pm, Guillermo Rauch <rau...@gmail.com> wrote:
> Just pull the latest tag =]
>
>
>
>
>
>
>
>
>
> On Mon, Jan 10, 2011 at 2:27 PM, Allen Lee <Allen....@asu.edu> wrote:
> > Nevermind, I just realized this is for socket.io-node and not the
> > client socket.io - doh!
>
Guillermo Rauch
Jan 10, 2011, 7:58:17 PM1/10/11
to sock...@googlegroups.com
Try pulling the very latest one.
Peter
Jan 10, 2011, 9:07:27 PM1/10/11
to Socket.IO
Guillermo, it's significantly better. Thank you. I still get
disconnects about once every 2-3 minutes, but it's definitely usable.
disconnects about once every 2-3 minutes, but it's definitely usable.
Guillermo Rauch
Jan 10, 2011, 9:31:34 PM1/10/11
to sock...@googlegroups.com
Is this on mobile only ? Can you try incrasing the closeTimeout ?
Peter
Jan 11, 2011, 1:46:26 AM1/11/11
to Socket.IO
Substantially better with a closeTimeout. I used 10s timeout and it
did not disconnect after using it for 5-10 minutes over wifi. I
switched to a 2G Edge network and it disconnected after a couple of
minutes. I think I would use 20-30s on an Edge network.
Maybe there can be some default high closeTimeout value if it detects
a mobile device... it's hardly usable otherwise
Thanks for all your help
On Jan 10, 6:31 pm, Guillermo Rauch <rau...@gmail.com> wrote:
> Is this on mobile only ? Can you try incrasing the closeTimeout ?
>
did not disconnect after using it for 5-10 minutes over wifi. I
switched to a 2G Edge network and it disconnected after a couple of
minutes. I think I would use 20-30s on an Edge network.
Maybe there can be some default high closeTimeout value if it detects
a mobile device... it's hardly usable otherwise
Thanks for all your help
On Jan 10, 6:31 pm, Guillermo Rauch <rau...@gmail.com> wrote:
> Is this on mobile only ? Can you try incrasing the closeTimeout ?
>
Arnout Kazemier
Jan 11, 2011, 2:13:28 AM1/11/11
to sock...@googlegroups.com
It's one thing to detect for a mobile device, but it would be harder to check if a 3G connection is used. Atleast if we want a higher timeout there
~ Arnout Kazemier
Reply all
Reply to author
Forward
0 new messages