[OTR-users] (no subject)

[Michael F. Wright]

> -----Original Message-----
> From: Ian Goldberg [mailto:ian at cypherpunks.ca] 
> Sent: Tuesday, March 22, 2005 10:42 AM
> To: michael.wright at janusresearch.com
> Cc: otr-users at lists.cypherpunks.ca
> Subject: Re: [OTR-users] (no subject)
> 
> On Tue, Mar 22, 2005 at 10:19:11AM -0500, Michael F. Wright wrote:
> > Hello!
> > 
> > I've been using OTR and GAIM for some time with great success. 
> > However, when attempting to use the OTR proxy with the AIM protocol 
> > under Trillian, it fails.
> > 
> > I installed the binary for Win32 located at:
> > http://www.cypherpunks.ca/otr/binaries/windows/gaim-otr-2.0.1.exe
> 
> [I assume you mean the otrproxy.exe, not the gaim-otr.exe.]


Yes. I apologize for this oversight. I was in a hurry when I posted. The
link I downloaded is at:
http://www.cypherpunks.ca/otr/binaries/windows/otrproxy-0.2.0-2.exe

:)

> 
> > One thing I noticed that's different bahavior than is expected from 
> > the README is that, when I start otrproxy, there is no output 
> > whatsoever. The output in the README that contains the 
> copyright statement, version, etc.
> > does not appear until after I CLOSE the UI for otrproxy, 
> which kills 
> > the process.
> 
> Aha!  We had gotten another report of "no output on Win32", 
> but this report of "doesn't show up until the UI closes" 
> suggests that Win32 needs fflush(stdout) more than other 
> platforms.  We'll fix this in CVS.
> 
> Thanks!


No problem.


> 
> > When I start it up and try to connect to AIM via Trillian, it never 
> > does connect. I have tried both the SOCKS5 and the HTTP proxies on 
> > ports 1080 and 8080 respectively.
> > 
> > Interestingly enough, when I run netstat after starting 
> otrproxy, I do 
> > see my system listening on those ports.
> > 
> > Any thoughts?
> 
> We've also had other reports of this happening 
> (intermittently; you should try again later) with Trillian.  
> It turns out there's about to be non-trivial changes to the 
> networking bits of the proxy, so we'll make sure to 
> well-instrument them, so that we can figure out what's going on.
> 
> One theory is that, on occasion, the AIM server just takes 
> too long to respond, and the TCP connect times out.  Somehow 
> this isn't being reported back to Trillian (the SOCKS5 
> connection should close if this happens), and Trillian never 
> retries it (which it does in the direct-connect case).
> 
> Thanks again for the bug reports!


I've used an ssh proxy in the past to circumvent firewalls and the like. By
setting up a dynamic port-forwarding ssh tunnel, I've successfully connected
to AIM through a SOCKS4 proxy, which the ssh tunnel builds. So, I know that
other proxies work in a timely fashion from this PC.

If you need any other info from me, let me know! I'm looking forward to more
great code from the OTR team!

> 
>    - Ian
>