tinc - Feb 2014 - tinc 1.1pre10, issues with Windows client

I compiled tinc 1.1pre10 for all my linux machines, just dropped in the 
new binaries, and restarted tinc. With 1.1pre10, all my linux machines 
are running and communicating just fine with each other.
I have one Windows machine on which I run tinc.  It sits behind a 
firewall (nat) and no longer will connect with the rest of my network.  
What I did was, remove the tinc service, uninstalled tinc 1.1pre9, 
installed 1.1pre10, made sure my exisiting config was intact, and 
reinstalled the service.

Here are some debug logs from one linux machine and my windows machine

tincd 1.1pre10 (Feb 11 2014 08:12:33) starting, debug level 5
/dev/net/tun is a Linux tun/tap device (tun mode)
Listening on 0.0.0.0 port 655
Listening on :: port 655
Executing script tinc-up
Ready
Connection from work.ip.addres port 1255
Sending ID to <unknown> (work.ip.addres port 1255): 0 home 17.3
Sending 12 bytes of metadata to <unknown> (work.ip.addres port 1255)
Metadata socket read error for <unknown> (work.ip.addres port 1255): 
Connection reset by peer


2014-02-11 10:16:34 tinc[3636]: tinc.vpn service installed
2014-02-11 10:16:34 tinc[3636]: tinc.vpn service started
2014-02-11 10:16:34 tinc.vpn[6200]: tincd 1.1pre10 (Feb  7 2014 
22:45:15) starting, debug level 5
2014-02-11 10:16:34 tinc.vpn[6200]: Tap reader running
2014-02-11 10:16:34 tinc.vpn[6200]: 
{40D064C7-74B2-4C8D-88CE-42C79B149E8A} (tinc) is a Windows tap device
2014-02-11 10:16:34 tinc.vpn[6200]: Listening on :: port 655
2014-02-11 10:16:34 tinc.vpn[6200]: Listening on 0.0.0.0 port 655
2014-02-11 10:16:34 tinc.vpn[6200]: Executing script tinc-up
2014-02-11 10:16:34 tinc.vpn[6200]: Ready
2014-02-11 10:16:34 tinc.vpn[6200]: Trying to connect to home 
(home.ip.address port 655)
2014-02-11 10:16:34 tinc.vpn[6200]: Broadcasting packet of 150 bytes 
from work (MYSELF port 655)
2014-02-11 10:16:34 tinc.vpn[6200]: Connected to home (home.ip.address 
port 655)
2014-02-11 10:16:34 tinc.vpn[6200]: Sending ID to home (home.ip.address 
port 655): 0 work 17.3
2014-02-11 10:16:34 tinc.vpn[6200]: Sending 12 bytes of metadata to home 
(home.ip.address port 655)
2014-02-11 10:16:34 tinc.vpn[6200]: Connection closed by home 
(home.ip.address port 655)
2014-02-11 10:16:34 tinc.vpn[6200]: Closing connection with home 
(home.ip.address port 655)
2014-02-11 10:16:34 tinc.vpn[6200]: Could not set up a meta connection 
to home

tinc.conf

Name = work
Interface = tinc
DeviceType = tun
Mode = router
ConnectTo = home (windows machine only)


All my hosts configs are this format.

-----BEGIN RSA PUBLIC KEY-----
-----END RSA PUBLIC KEY-----
ECDSAPublicKey = removed
Address = homeipaddress.com
Subnet = 192.168.10.0/24
Compression = 10
Cipher = AES-256-CBC

On Tue, Mar 11, 2014 at 02:50:14PM -0600, Lance Fredrickson wrote:
> >>When reverting all linux nodes and the windows node back to
1.1pre9, then
> >>there is no issue, all while using the  identical config files.
> >What errors did tinc log when the Windows node failed to connect?
[...]
> I started a separate email thread describing the issue, but it must
> have been passed over, or I sent it wrong.  For brevity here's a
> link I googled of the email.
> http://permalink.gmane.org/gmane.network.tinc.user/2171
Sorry, I did indeed pass over that one. It is strange to see both sides
claiming that the other side closed the connection. I don't know (yet) what
could have gone wrong in your situation.

-- 
Met vriendelijke groet / with kind regards,
     Guus Sliepen <guus at tinc-vpn.org>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL:
<http://www.tinc-vpn.org/pipermail/tinc/attachments/20140311/95abf6b7/attachment.sig>