From anthony at interlink.com.au  Sun Jan  9 13:08:23 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Sun, 9 Jan 2005 23:08:23 +1100
Subject: [Shtoom] UPnP, improved STUN support are checked in
Message-ID: <200501092308.25070.anthony@interlink.com.au>

I've just checked in support for UPnP and new STUN code,
from the improved-stun branch. This should make things much,
much better than they were before. 

Note that as part of this, the minimum required version of 
Twisted is now 1.3. 

This also requires the 'zope.interface' package, available from 
http://zope.org/Products/ZopeInterface

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From anthony at interlink.com.au  Sun Jan  9 13:11:04 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Sun, 9 Jan 2005 23:11:04 +1100
Subject: [Shtoom] Re: UPnP, improved STUN support are checked in
In-Reply-To: <200501092308.25070.anthony@interlink.com.au>
References: <200501092308.25070.anthony@interlink.com.au>
Message-ID: <200501092311.05139.anthony@interlink.com.au>

On Sunday 09 January 2005 23:08, Anthony Baxter wrote:
> I've just checked in support for UPnP and new STUN code,
> from the improved-stun branch. This should make things much,
> much better than they were before.

I forgot to add - there's a new 'shtoominfo.py' script in the scripts
directory. This should output a bunch of useful information. Please
include it's output in any bug reports.

Anthony

-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From anthony at interlink.com.au  Wed Jan 12 12:50:55 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 12 Jan 2005 22:50:55 +1100
Subject: [Shtoom] Re: [Divmod-dev] integrating Shtoom
In-Reply-To: <FF32CCF2-62EE-11D9-B434-000D933B4B48@aya.yale.edu>
References: <FF32CCF2-62EE-11D9-B434-000D933B4B48@aya.yale.edu>
Message-ID: <200501122250.57734.anthony@interlink.com.au>

[followups should probably go to the shtoom at python.org list]

On Monday 10 January 2005 21:04, Eric Ries wrote:
> Hey there,
>
> I'm the author of a program called IMVU, which is an IM add-on for
> existing IM services. It's pretty fun, you can read about it here:
> www.imvu.com
>
> IMVU is written in Python/C++ and I thought it might be interesting to
> try embedding Shtoom as part of it.

Sure - one of the goals of shtoom is that is should be possible to embed
it in other applications. What sort of things did you have in mind?

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From anthony at interlink.com.au  Wed Jan 12 17:31:42 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Thu, 13 Jan 2005 03:31:42 +1100
Subject: [Shtoom] CocoaShtoom snapshots now available
Message-ID: <200501130331.42768.anthony@interlink.com.au>

Thanks to the heroic efforts of a cast of thousands [1]
there's now a CocoaShtoom application available. It's very
much a snapshot/work-in-progress thing, but it seems to
work for me. If you want to try it out and let me know if 
it works/doesn't work for you, that'd be great!

They're available from 
http://www.interlink.com.au/anthony/tech/shtoom/

This really is a first cut - it's quite possible that while it works
for me, it will utterly fail for you in very bad ways. Nonetheless,
if you want to try it out, that'd be great! 

It's built with GSM and ulaw audio - I'll look at getting the Speex
stuff built at some point.

If you have a Mac (OSX 10.3 required) and want to try
it out, please do so and let us know how it works for you
(either on this list, or to #shtoom on irc.freenode.net).
You can try calling sip:613 at fwd.pulver.com for a simple
echo-test.

And yes, windows executables are the next cab off the rank.

Anthony

[1] well, ok - Bob and Donovan did the heavy lifting, and a
magnificent job of it they did, too.
-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From eric.ries at aya.yale.edu  Wed Jan 12 20:48:52 2005
From: eric.ries at aya.yale.edu (Eric Ries)
Date: Wed, 12 Jan 2005 11:48:52 -0800
Subject: [Shtoom] Re: [Divmod-dev] integrating Shtoom
In-Reply-To: <200501122250.57734.anthony@interlink.com.au>
References: <FF32CCF2-62EE-11D9-B434-000D933B4B48@aya.yale.edu>
	<200501122250.57734.anthony@interlink.com.au>
Message-ID: <FCB00002-64D2-11D9-B434-000D933B4B48@aya.yale.edu>

IMVU is an IM add-on that provides an immersive 3D environment for 
chat. Several of our users have suggested adding voice-chat 
capabilities, which I think is a cool idea. So I've been looking around 
for something that'd be interesting to embed. I thought it might be an 
opportunity to bring Shtoom to a pretty big audience.

We often hire contributors  who work on open source projects as 
contractors to help us. Naturally, we ask that any improvements they 
make to the project be contributed back under a free software license. 
If you're interested, please let me know.

Eric

On Jan 12, 2005, at 3:50 AM, Anthony Baxter wrote:

> [followups should probably go to the shtoom at python.org list]
>
> On Monday 10 January 2005 21:04, Eric Ries wrote:
>> Hey there,
>>
>> I'm the author of a program called IMVU, which is an IM add-on for
>> existing IM services. It's pretty fun, you can read about it here:
>> www.imvu.com
>>
>> IMVU is written in Python/C++ and I thought it might be interesting to
>> try embedding Shtoom as part of it.
>
> Sure - one of the goals of shtoom is that is should be possible to 
> embed
> it in other applications. What sort of things did you have in mind?
>
> Anthony
> -- 
> Anthony Baxter     <anthony at interlink.com.au>
> It's never too late to have a happy childhood.
>



From zooko at zooko.com  Wed Jan 19 21:49:44 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 19 Jan 2005 16:49:44 -0400
Subject: [Shtoom] bug report: Shtoom doesn't register on startup
Message-ID: <A615D59C-6A5B-11D9-B107-000A95E2A184@zooko.com>

Current SVN head doesn't register with a SIP server on startup.  Here  
is a complete log of Shtoom starting and not registering.

Regards,

Zooko


2005/01/19 16:45 AST [-] Log opened.
2005-01-19 16:45:32.001 Shtoom[7619] ShtoomAudioMonitor setDelegate
2005-01-19 16:45:32.057 Shtoom[7619] ShtoomAudioMonitor awakeFromNib
2005-01-19 16:45:32.058 Shtoom[7619] ShtoomAudioMonitor setSampleRate
2005-01-19 16:45:32.435 Shtoom[7619]  
MTAudioDeviceBrowser:selectedDeviceDidChange:
2005-01-19 16:45:32.436 Shtoom[7619] setDeviceOpen:
2005-01-19 16:45:32.436 Shtoom[7619] Stopping devices.
2005-01-19 16:45:32.442 Shtoom[7619]  
MTAudioDeviceBrowser:selectedDeviceDidChange:
2005-01-19 16:45:32.443 Shtoom[7619] setDeviceOpen:
2005-01-19 16:45:32.443 Shtoom[7619] Stopping devices.
2005-01-19 16:45:32.444 Shtoom[7619] setDeviceOpen:
2005-01-19 16:45:32.444 Shtoom[7619] Stopping devices.
2005-01-19 16:45:32.598 Shtoom[7619] Application did finish launching.
2005-01-19 16:45:32.600 Shtoom[7619] Audio Monitor =  
<ShtoomAudioMonitor: 0xb68990>
2005/01/19 16:45 AST [-] <class 'shtoom.sip.SipProtocol'> starting on  
5066
2005/01/19 16:45 AST [-] Starting protocol <shtoom.sip.SipProtocol  
object at 0x1fe20b0>
2005/01/19 16:45 AST [-]  
/Users/zooko/playground/shtoom/svnworld/trunk/CocoaShtoom/dist/ 
Shtoom.app/Contents/Resources/Python/site-packages/shtoom/app/base.py: 
54: exceptions.DeprecationWarning: IPv4Address.__getitem__ is  
deprecated.  Use attributes instead.
2005/01/19 16:45 AST [app] sip listener installed on 5066
2005/01/19 16:45 AST [-] <class 'shtoom.upnp.UPnPProtocol'> starting on  
1900
2005/01/19 16:45 AST [-] Starting protocol <shtoom.upnp.UPnPProtocol  
object at 0x1fa3cd0>
2005/01/19 16:45 AST [-] <class 'shtoom.stun._DetectSTUNProt'> starting  
on 0
2005/01/19 16:45 AST [-] Starting protocol <shtoom.stun._DetectSTUNProt  
object at 0x1fc3670>
2005/01/19 16:45 AST [-]  
/Users/zooko/playground/shtoom/svnworld/trunk/CocoaShtoom/dist/ 
Shtoom.app/Contents/Resources/Python/site-packages/nevow/static.py:25:  
exceptions.DeprecationWarning: twisted.protocols.http is DEPRECATED.  
See http://twistedmatrix.com/projects/web.
2005/01/19 16:45 AST [-] NamedAnyError: Name  
twisted.web.woven.guard.GuardSession not found.
2005/01/19 16:45 AST [-] NamedAnyError: Name  
twisted.web.woven.guard.GuardSession not found.
2005/01/19 16:45 AST [-] nevow.appserver.NevowSite starting on 0
2005/01/19 16:45 AST [-] Starting factory <nevow.appserver.NevowSite  
instance at 0x6e9cd50>
2005-01-19 16:45:34.058 Shtoom[7619] Preferences server listening on  
port 53237.
2005/01/19 16:45 AST [sip] no outstanding registrations, registering
2005/01/19 16:45 AST [-]  
/Users/zooko/playground/shtoom/svnworld/trunk/CocoaShtoom/dist/ 
Shtoom.app/Contents/Resources/Python/site-packages/shtoom/sip.py:294:  
exceptions.DeprecationWarning: use listenUDP and then  
transport.connect().
2005/01/19 16:45 AST [-]  
twisted.internet.protocol.ConnectedDatagramProtocol starting on 0
2005/01/19 16:45 AST [-] Starting protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x6eb4198>
2005/01/19 16:45 AST [-]  
/Users/zooko/playground/shtoom/svnworld/trunk/CocoaShtoom/dist/ 
Shtoom.app/Contents/Resources/Python/site-packages/shtoom/sip.py:299:  
exceptions.DeprecationWarning: IPv4Address.__getitem__ is deprecated.   
Use attributes instead.
2005/01/19 16:45 AST [-]  
/Users/zooko/playground/shtoom/svnworld/trunk/CocoaShtoom/dist/ 
Shtoom.app/Contents/Resources/Python/site-packages/shtoom/sip.py:300:  
exceptions.DeprecationWarning: IPv4Address.__getitem__ is deprecated.   
Use attributes instead.
2005/01/19 16:45 AST [sip] discovered local address ('192.168.1.110',  
5066), remote ('divmod.com', 5060)
2005/01/19 16:45 AST [sip] NAT policy says yes, we're transitting a NAT
2005/01/19 16:45 AST [-]  
twisted.internet.protocol.ConnectedDatagramProtocol starting on 0
2005/01/19 16:45 AST [-] Starting protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x6eb4580>
2005/01/19 16:45 AST [-] (Port 0 Closed)
2005/01/19 16:45 AST [-] Stopping protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x6eb4198>
2005/01/19 16:45 AST [stun] got STUN response from ('64.69.76.23', 3478)
2005/01/19 16:45 AST [-] (Port 0 Closed)
2005/01/19 16:45 AST [-] Stopping protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x6eb4580>
2005/01/19 16:45 AST [UPnP] got a response from ('192.168.1.1', 1900),  
status 200
2005/01/19 16:45 AST [UPnP] got a response from ('192.168.1.1', 1900),  
status 200
2005/01/19 16:45 AST [UPnP] got a response from ('192.168.1.1', 1900),  
status 200
2005/01/19 16:45 AST [UPnP] you're behind a Linksys Inc.
2005/01/19 16:45 AST [UPnP] 'upnp response has no urlbase?!?'
2005/01/19 16:45 AST [UPnP] you're behind a Linksys Inc.
2005/01/19 16:45 AST [UPnP] 'upnp response has no urlbase?!?'
2005/01/19 16:45 AST [UPnP] you're behind a Linksys Inc.
2005/01/19 16:45 AST [UPnP] 'upnp response has no urlbase?!?'
2005/01/19 16:45 AST [stun] got STUN response from ('64.69.76.24', 3478)
2005/01/19 16:45 AST [-] (Port 0 Closed)
2005/01/19 16:45 AST [-] Stopping protocol <shtoom.stun._DetectSTUNProt  
object at 0x1fc3670>



From zooko at zooko.com  Wed Jan 19 21:58:57 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 19 Jan 2005 16:58:57 -0400
Subject: [Shtoom] bug report followup: it works when not behind NAT
Message-ID: <EFFD7E3F-6A5C-11D9-B175-000A95E2A184@zooko.com>

I plugged into the Real Internet and tried again, and it registered.   
Here is the current shtoominfo when plugged into the Real Internet.

HACK draco:~/playground/shtoom/svnworld/trunk$  
./shtoom/scripts/shtoominfo.py
Shtoom, version 0.3alpha0
Using python version 2.3.0
Using twisted version SVN-trunk
Running on Darwin Power Macintosh 7.7.0
Available audio interfaces: fileaudio
Available codecs: mulaw
Local IP address: 24.215.68.128
No UPnP-capable device discovered
STUN says NAT type: FullCone
And the mapper we'd use is: <shtoom.stun.STUNMapper instance at  
0x610648>

Here is the log:

2005/01/19 16:55 AST [-] Log opened.
2005-01-19 16:55:35.890 Shtoom[7752] ShtoomAudioMonitor setDelegate
2005-01-19 16:55:35.989 Shtoom[7752] ShtoomAudioMonitor awakeFromNib
2005-01-19 16:55:35.990 Shtoom[7752] ShtoomAudioMonitor setSampleRate
2005-01-19 16:55:36.136 Shtoom[7752]  
MTAudioDeviceBrowser:selectedDeviceDidChange:
2005-01-19 16:55:36.136 Shtoom[7752] setDeviceOpen:
2005-01-19 16:55:36.137 Shtoom[7752] Stopping devices.
2005-01-19 16:55:36.143 Shtoom[7752]  
MTAudioDeviceBrowser:selectedDeviceDidChange:
2005-01-19 16:55:36.143 Shtoom[7752] setDeviceOpen:
2005-01-19 16:55:36.143 Shtoom[7752] Stopping devices.
2005-01-19 16:55:36.144 Shtoom[7752] setDeviceOpen:
2005-01-19 16:55:36.144 Shtoom[7752] Stopping devices.
2005-01-19 16:55:36.371 Shtoom[7752] Application did finish launching.
2005-01-19 16:55:36.373 Shtoom[7752] Audio Monitor =  
<ShtoomAudioMonitor: 0xb9cde0>
2005/01/19 16:55 AST [-] Zfone devel revision zfone v365, platform:  
Darwin-7.7.0-Power_Macintosh-powerpc-32bit
2005/01/19 16:55 AST [-] <class 'shtoom.sip.SipProtocol'> starting on  
5066
2005/01/19 16:55 AST [-] Starting protocol <shtoom.sip.SipProtocol  
object at 0xa5a990>
2005/01/19 16:55 AST [app] sip listener installed on 5066
2005/01/19 16:55 AST [-] <class 'shtoom.upnp.UPnPProtocol'> starting on  
1900
2005/01/19 16:55 AST [-] Starting protocol <shtoom.upnp.UPnPProtocol  
object at 0x6d120d0>
2005/01/19 16:55 AST [-] <class 'shtoom.stun._DetectSTUNProt'> starting  
on 0
2005/01/19 16:55 AST [-] Starting protocol <shtoom.stun._DetectSTUNProt  
object at 0x6d1ef70>
2005/01/19 16:55 AST [-] NamedAnyError: Name  
twisted.web.woven.guard.GuardSession not found.
2005/01/19 16:55 AST [-] NamedAnyError: Name  
twisted.web.woven.guard.GuardSession not found.
2005/01/19 16:55 AST [-] nevow.appserver.NevowSite starting on 0
2005/01/19 16:55 AST [-] Starting factory <nevow.appserver.NevowSite  
instance at 0x7042da0>
2005-01-19 16:55:38.798 Shtoom[7752] Preferences server listening on  
port 53419.
2005/01/19 16:55 AST [sip] no outstanding registrations, registering
2005/01/19 16:55 AST [-]  
twisted.internet.protocol.ConnectedDatagramProtocol starting on 0
2005/01/19 16:55 AST [-] Starting protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x70581e8>
2005/01/19 16:55 AST [sip] discovered local address ('24.215.68.128',  
5066), remote ('divmod.com', 5060)
2005/01/19 16:55 AST [-] Debug: register sent
	REGISTER sip:divmod.com:5060 SIP/2.0

	Via: SIP/2.0/UDP 24.215.68.128:5066;rport

	CSeq: 2697 REGISTER

	To: sip:zooko at divmod.com:5060

	From: sip:zooko at divmod.com:5060

	Expires: 900

	Call-ID: 400993727 at 24.215.68.128

	User-Agent: Shtoom/0.3alpha0

	Contact: <sip:zooko at 24.215.68.128:5066>

	Content-Length: 0

	

	
2005/01/19 16:55 AST [-]  
twisted.internet.protocol.ConnectedDatagramProtocol starting on 0
2005/01/19 16:55 AST [-] Starting protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x70585d0>
2005/01/19 16:55 AST [-] (Port 0 Closed)
2005/01/19 16:55 AST [-] Stopping protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x70581e8>
2005/01/19 16:55 AST [UPnP] UPnP discovery timed out
2005/01/19 16:55 AST [UPnP] no UPnP found!
2005/01/19 16:55 AST [-] (Port 0 Closed)
2005/01/19 16:55 AST [-] Stopping protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x70585d0>
2005/01/19 16:55 AST [stun] got STUN response from ('69.90.168.14',  
3478)
2005/01/19 16:55 AST [-] Debug: Got a SIP packet from  
198.49.126.192:5060
2005/01/19 16:55 AST [-] Debug: SIP PACKET
	SIP/2.0 401 Unauthorized

	Via: SIP/2.0/UDP 24.215.68.128:5066;received=24.215.68.128;rport=5066

	To: sip:zooko at divmod.com:5060

	From: sip:zooko at divmod.com:5060

	Call-ID: 400993727 at 24.215.68.128

	CSeq: 2697 REGISTER

	WWW-Authenticate: Digest  
nonce="10345170551339095801663502492",opaque="657028545",qop="auth",algo 
rithm="MD5",realm="divmod.com"

	

	
2005/01/19 16:55 AST [-] Debug: got SIP response 401: Unauthorized
2005/01/19 16:55 AST [-] Debug: got SIP response
	 SIP/2.0 401 Unauthorized

	Via: SIP/2.0/UDP 24.215.68.128:5066;received=24.215.68.128;rport=5066

	To: sip:zooko at divmod.com:5060

	From: sip:zooko at divmod.com:5060

	Call-ID: 400993727 at 24.215.68.128

	CSeq: 2697 REGISTER

	WWW-Authenticate: Digest  
nonce="10345170551339095801663502492",opaque="657028545",qop="auth",algo 
rithm="MD5",realm="divmod.com"

	

	
2005/01/19 16:55 AST [-] Debug: register sent
	REGISTER sip:divmod.com:5060 SIP/2.0

	Via: SIP/2.0/UDP 24.215.68.128:5066;rport

	CSeq: 2698 REGISTER

	To: sip:zooko at divmod.com:5060

	From: sip:zooko at divmod.com:5060

	Expires: 900

	Call-ID: 400993727 at 24.215.68.128

	Authorization: Digest username="zooko", realm="divmod.com",  
nonce="10345170551339095801663502492", uri="sip:divmod.com:5060",  
response="01acfb74354433ff7b7a2efd6a0d47f2", opaque="657028545",  
qop=auth, nc=00000001, cnonce="ed00"

	User-Agent: Shtoom/0.3alpha0

	Contact: <sip:zooko at 24.215.68.128:5066>

	Content-Length: 0

	

	
2005/01/19 16:55 AST [stun] got STUN response from ('69.90.168.13',  
3479)
2005/01/19 16:55 AST [nat] detectNAT got [(True, None), (True, <NatType  
FullCone>)]
2005/01/19 16:55 AST [nat] using STUN mapper
2005/01/19 16:55 AST [-] <class 'shtoom.stun._DetectSTUNProt'> starting  
on 0
2005/01/19 16:55 AST [-] Starting protocol <shtoom.stun._DetectSTUNProt  
object at 0x7055650>
2005/01/19 16:55 AST [-] (Port 0 Closed)
2005/01/19 16:55 AST [-] Stopping protocol <shtoom.stun._DetectSTUNProt  
object at 0x6d1ef70>
2005/01/19 16:55 AST [-]  
twisted.internet.protocol.ConnectedDatagramProtocol starting on 0
2005/01/19 16:55 AST [-] Starting protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x6d32c38>
2005/01/19 16:55 AST [stun] got STUN response from ('69.90.168.14',  
3478)
2005/01/19 16:55 AST [-] (Port 0 Closed)
2005/01/19 16:55 AST [-] Stopping protocol  
<twisted.internet.protocol.ConnectedDatagramProtocol instance at  
0x6d32c38>
2005/01/19 16:55 AST [-] Debug: Got a SIP packet from  
198.49.126.192:5060
2005/01/19 16:55 AST [-] Debug: SIP PACKET
	SIP/2.0 200 OK

	Via: SIP/2.0/UDP 24.215.68.128:5066;received=24.215.68.128;rport=5066

	To: sip:zooko at divmod.com:5060

	From: sip:zooko at divmod.com:5060

	Call-ID: 400993727 at 24.215.68.128

	CSeq: 2698 REGISTER

	Contact: sip:zooko at 24.215.68.128:5066

	Expires: 900

	Content-length: 0

	

	
2005/01/19 16:55 AST [-] Debug: got SIP response 200: OK
2005/01/19 16:55 AST [-] Debug: got SIP response
	 SIP/2.0 200 OK

	Via: SIP/2.0/UDP 24.215.68.128:5066;received=24.215.68.128;rport=5066

	To: sip:zooko at divmod.com:5060

	From: sip:zooko at divmod.com:5060

	Call-ID: 400993727 at 24.215.68.128

	CSeq: 2698 REGISTER

	Contact: sip:zooko at 24.215.68.128:5066

	Expires: 900

	Content-Length: 0

	

	
2005/01/19 16:55 AST [stun] got STUN response from ('69.90.168.13',  
3479)
2005/01/19 16:55 AST [-] (Port 0 Closed)
2005/01/19 16:55 AST [-] Stopping protocol <shtoom.stun._DetectSTUNProt  
object at 0x7055650>
2005/01/19 16:55 AST [stun] got STUN response from ('69.90.168.14',  
3478)
2005/01/19 16:55 AST [sip] late stun packet, ignoring
2005/01/19 16:55 AST [sip] late stun packet, ignoring
2005/01/19 16:55 AST [-] Annnnd we're dying
2005/01/19 16:55 AST [-] stopping SIP, and unmapping it (True)
2005/01/19 16:55 AST [-] unmapping sip (<shtoom.stun.STUNMapper  
instance at 0x705b3f0>)
2005/01/19 16:55 AST [nevow.appserver.NevowSite] (Port 0 Closed)
2005/01/19 16:55 AST [nevow.appserver.NevowSite] Stopping factory  
<nevow.appserver.NevowSite instance at 0x7042da0>
2005/01/19 16:55 AST [shtoom.upnp.UPnPProtocol (UDP)] (Port 1900 Closed)
2005/01/19 16:55 AST [shtoom.upnp.UPnPProtocol (UDP)] Stopping protocol  
<shtoom.upnp.UPnPProtocol object at 0x6d120d0>
	



From zooko at zooko.com  Wed Jan 19 22:03:29 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 19 Jan 2005 17:03:29 -0400
Subject: [Shtoom] followup: shtoominfo from NAT
Message-ID: <919A363C-6A5D-11D9-B175-000A95E2A184@zooko.com>

Here is the result of shtoominfo,py when behind my Linksys wrt54g NAT.  
Note the time to complete.  The Linksys has firmware revision 3.01.3.  
Earlier firmwares in this box reached into SIP packets and mangled the 
RTP port number  -- it added 2 to the media port number.

Err..  Actually shtoominfo.py is taking too long to complete so I'm 
going to send this mail without it.  The progress so far is:

HACK draco:~/playground/shtoom/svnworld/trunk$ time 
./shtoom/scripts/shtoominfo.py
'upnp response has no urlbase?!?'
'upnp response has no urlbase?!?'
'upnp response has no urlbase?!?'



From anthony at interlink.com.au  Thu Jan 20 11:22:56 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Thu, 20 Jan 2005 21:22:56 +1100
Subject: [Shtoom] bug report: Shtoom doesn't register on startup
In-Reply-To: <A615D59C-6A5B-11D9-B107-000A95E2A184@zooko.com>
References: <A615D59C-6A5B-11D9-B107-000A95E2A184@zooko.com>
Message-ID: <200501202122.57562.anthony@interlink.com.au>

On Thursday 20 January 2005 07:49, Zooko O'Whielacronx wrote:
> Current SVN head doesn't register with a SIP server on startup.  Here
> is a complete log of Shtoom starting and not registering.

I made some changes to the NAT code yesterday to attempt to better
handle Symmetric NATs. As far as I can test it, it works fine - you might
try backing out to r1145 and see if that makes it better.

(Aside: Testing this NAT stuff is a FRICKING NIGHTMARE!!!!)

-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From liste at zerozone.it  Fri Jan 21 12:48:27 2005
From: liste at zerozone.it (O-Zone)
Date: Fri, 21 Jan 2005 12:48:27 +0100
Subject: [Shtoom] Data FLow
Message-ID: <200501211248.27340@X-Blurb>

Hi Shtoom developers,
i'm working on a SIP client (in Python and WX) heavily based on Shtoom. I 
decide to do this for some reasons:

1) Shtoom is not very user-friedly
2) Shtoom have a lot of GUIs that can confuse dummy users
3) Shtoom isn't very beatiful (GUI)

so my SIP client (UniversalSIPClient) try to resolve this tree points to have 
a professiona GPL product to share. At this point REGISTRATION works and also 
CALL ACKs but i need to do something abount codecs and audio managment. 
Because i want to do a portable client, i decide to use pyMedia package for 
audio (perhaps not the best solution available !) but this don't have CODECS 
support. I'm a bit confused also on AUDIO FLOW: what happen after user call 
acceptCall() function ? Where i get the audio data ?

Thanks! Oz  

-- 
------
O-Zone ! www.zerozone.it


From anthony at interlink.com.au  Sat Jan 22 05:11:44 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Sat, 22 Jan 2005 15:11:44 +1100
Subject: [Shtoom] Data FLow
In-Reply-To: <200501211248.27340@X-Blurb>
References: <200501211248.27340@X-Blurb>
Message-ID: <200501221511.45340.anthony@interlink.com.au>

On Friday 21 January 2005 22:48, O-Zone wrote:
> At this point REGISTRATION works
> and also CALL ACKs but i need to do something abount codecs and audio
> managment. Because i want to do a portable client, i decide to use pyMedia
> package for audio (perhaps not the best solution available !) but this
> don't have CODECS support. I'm a bit confused also on AUDIO FLOW: what
> happen after user call acceptCall() function ? Where i get the audio data ?

I doubt the pymedia package is suitable for this. The codec support is in 
shtoom.audio.converters.

Note that in the next few weeks I'll be making some fairly fundamental
refactorings that will make the process of plugging in, say, a new GUI 
a much much more trivial exercise. Stay tuned.

Anthony

-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From zooko at zooko.com  Sat Jan 22 17:19:18 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 22 Jan 2005 12:19:18 -0400
Subject: [Shtoom] bug report: test_stunmapper et al. sporadic failure
Message-ID: <5E11CDD6-6C91-11D9-9811-000A95E2A184@zooko.com>

I thought that test_stunmapper et al. were failing due to their timeout 
being too short.  I thought this because they failed, I doubled the 
timeout, and then they passed.  However, right now I've set the timeout 
on test_stunmapper() up to 64 seconds, and it still fails about 1 out 
of 3 runs.

Regards,

Zooko



From zooko at zooko.com  Sat Jan 22 17:28:07 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 22 Jan 2005 12:28:07 -0400
Subject: [Shtoom] Re: bug report: test_stunmapper et al. sporadic failure
In-Reply-To: <5E11CDD6-6C91-11D9-9811-000A95E2A184@zooko.com>
References: <5E11CDD6-6C91-11D9-9811-000A95E2A184@zooko.com>
Message-ID: <99625A7E-6C92-11D9-9811-000A95E2A184@zooko.com>

Whoops -- I sent that msg before I meant to.  Actually, test_stunmapper 
has never yet failed with timeout 64.  I've run it five times.

--Z



From zooko at zooko.com  Sat Jan 22 17:42:15 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 22 Jan 2005 12:42:15 -0400
Subject: [Shtoom] Re: bug report: test_stunmapper et al. sporadic failure
In-Reply-To: <99625A7E-6C92-11D9-9811-000A95E2A184@zooko.com>
References: <5E11CDD6-6C91-11D9-9811-000A95E2A184@zooko.com>
	<99625A7E-6C92-11D9-9811-000A95E2A184@zooko.com>
Message-ID: <92A2E1EE-6C94-11D9-9811-000A95E2A184@zooko.com>

> Whoops -- I sent that msg before I meant to.  Actually, 
> test_stunmapper has never yet failed with timeout 64.  I've run it 
> five times.

But it *does* sporadically fail around 1 out of 3 times with timeout == 
32.

--Z



From zooko at zooko.com  Sat Jan 22 17:51:22 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 22 Jan 2005 12:51:22 -0400
Subject: [Shtoom] Re: bug report: test_stunmapper et al. sporadic failure
In-Reply-To: <92A2E1EE-6C94-11D9-9811-000A95E2A184@zooko.com>
References: <5E11CDD6-6C91-11D9-9811-000A95E2A184@zooko.com>
	<99625A7E-6C92-11D9-9811-000A95E2A184@zooko.com>
	<92A2E1EE-6C94-11D9-9811-000A95E2A184@zooko.com>
Message-ID: <D868D82E-6C95-11D9-9811-000A95E2A184@zooko.com>

Okay, now it failed once with timeout == 64.



From seberino at spawar.navy.mil  Sat Jan 22 19:39:37 2005
From: seberino at spawar.navy.mil (seberino at spawar.navy.mil)
Date: Sat, 22 Jan 2005 10:39:37 -0800
Subject: [Shtoom] Shtoom/SIP working (thru NAT!) ????
In-Reply-To: <5E11CDD6-6C91-11D9-9811-000A95E2A184@zooko.com>
References: <5E11CDD6-6C91-11D9-9811-000A95E2A184@zooko.com>
Message-ID: <20050122183937.GA10154@spawar.navy.mil>

Is Shtoom working behind my NATing firewall??  Looks like perhaps yes...


Here is my .shtoomrc and my shtoomphone.py output...

[shtoom]
audio=oss
ui=qt

[identity]
username=seberino

[register]
register_uri=sip:switch-2.nufone.net
register_authuser=seberino
register_user=seberino
register_authpasswd=XXXXX


===================================================

2005/01/22 10:32 PST [-] Log opened.
2005/01/22 10:32 PST [-] <class 'shtoom.sip.SipProtocol'> starting on
5060
2005/01/22 10:32 PST [-] Starting protocol <shtoom.sip.SipProtocol
object at 0xb7b9b82c>
2005/01/22 10:32 PST [app] sip listener installed on 5060
2005/01/22 10:32 PST [-] no outstanding registrations, registering
2005/01/22 10:32 PST [-]
twisted.internet.protocol.ConnectedDatagramProtocol starting on 0
2005/01/22 10:32 PST [-] Starting protocol
<twisted.internet.protocol.ConnectedDatagramProtocol instance at
0xb7ac3a8c>
2005/01/22 10:32 PST [sip] discovered local address ('192.168.100.3',
5060), remote ('switch-2.nufone.net', 5060)
2005/01/22 10:32 PST [sip] stun policy says yes, use STUN
2005/01/22 10:32 PST [-] (Port 0 Closed)
2005/01/22 10:32 PST [-] Stopping protocol
<twisted.internet.protocol.ConnectedDatagramProtocol instance at
0xb7ac3a8c>
2005/01/22 10:32 PST [stun] got STUN response from ('64.69.76.23',
3478)
2005/01/22 10:32 PST [sip] according to STUN, local address is
66.159.194.51:5060
2005/01/22 10:32 PST [ui] register sent
 REGISTER sip:seberino at switch-2.nufone.net SIP/2.0

 Via: SIP/2.0/UDP 66.159.194.51:5060

 CSeq: 3363 REGISTER

 To: sip:seberino at switch-2.nufone.net

 From: sip:seberino at switch-2.nufone.net

 Expires: 900

 Call-ID: 401507385 at 66.159.194.51

 User-Agent: Shtoom/0.3alpha0

 Contact: <sip:seberino at 66.159.194.51:5060>

 Content-Length: 0




2005/01/22 10:32 PST [ui] Got a SIP packet from 198.22.67.70:5060
2005/01/22 10:32 PST [ui] SIP PACKET
 SIP/2.0 100 Trying

 Via: SIP/2.0/UDP 66.159.194.51:5060;received=66.159.194.51;rport=5060

 From: sip:seberino at switch-2.nufone.net

 To: sip:seberino at switch-2.nufone.net;tag=as4a099353

 Call-ID: 401507385 at 66.159.194.51

 CSeq: 3363 REGISTER

 User-Agent: Asterisk PBX

 Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER

 Contact: <sip:seberino at 198.22.67.70>

 Content-Length: 0




2005/01/22 10:32 PST [ui] got SIP response 100: Trying
2005/01/22 10:32 PST [ui] got SIP response
  SIP/2.0 100 Trying

 Via: SIP/2.0/UDP 66.159.194.51:5060;received=66.159.194.51;rport=5060

 From: sip:seberino at switch-2.nufone.net

 To: sip:seberino at switch-2.nufone.net;tag=as4a099353

 Call-ID: 401507385 at 66.159.194.51

 CSeq: 3363 REGISTER

 User-Agent: Asterisk PBX

 Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER

 Contact: <sip:seberino at 198.22.67.70>

 Content-Length: 0




2005/01/22 10:32 PST [ui] Got a SIP packet from 198.22.67.70:5060
2005/01/22 10:32 PST [ui] SIP PACKET
 SIP/2.0 401 Unauthorized

 Via: SIP/2.0/UDP 66.159.194.51:5060;received=66.159.194.51;rport=5060

 From: sip:seberino at switch-2.nufone.net

 To: sip:seberino at switch-2.nufone.net;tag=as4a099353

 Call-ID: 401507385 at 66.159.194.51

 CSeq: 3363 REGISTER

 User-Agent: Asterisk PBX

 Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER

 Contact: <sip:seberino at 198.22.67.70>

 WWW-Authenticate: Digest realm="asterisk", nonce="2f70b9ac"

 Content-Length: 0




2005/01/22 10:32 PST [ui] got SIP response 401: Unauthorized
2005/01/22 10:32 PST [ui] got SIP response
  SIP/2.0 401 Unauthorized

 Via: SIP/2.0/UDP 66.159.194.51:5060;received=66.159.194.51;rport=5060

 From: sip:seberino at switch-2.nufone.net

 To: sip:seberino at switch-2.nufone.net;tag=as4a099353

 Call-ID: 401507385 at 66.159.194.51

 CSeq: 3363 REGISTER

 User-Agent: Asterisk PBX

 Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER

 Contact: <sip:seberino at 198.22.67.70>

 WWW-Authenticate: Digest realm="asterisk", nonce="2f70b9ac"

 Content-Length: 0

2005/01/22 10:32 PST [ui] register sent
 REGISTER sip:seberino at switch-2.nufone.net SIP/2.0

 Via: SIP/2.0/UDP 66.159.194.51:5060

 CSeq: 3364 REGISTER

 To: sip:seberino at switch-2.nufone.net

 From: sip:seberino at switch-2.nufone.net

 Expires: 900

 Call-ID: 401507385 at 66.159.194.51

 Authorization: Digest username="seberino", realm="asterisk",
 nonce="2f70b9ac", uri="sip:seberino at switch-2.nufone.net",
 response="513a840ef683366e7be5c06d3a1286ea"

 User-Agent: Shtoom/0.3alpha0

 Contact: <sip:seberino at 66.159.194.51:5060>

 Content-Length: 0




2005/01/22 10:32 PST [ui] Got a SIP packet from 198.22.67.70:5060
2005/01/22 10:32 PST [ui] SIP PACKET
 SIP/2.0 100 Trying

 Via: SIP/2.0/UDP 66.159.194.51:5060;received=66.159.194.51;rport=5060

 From: sip:seberino at switch-2.nufone.net

 To: sip:seberino at switch-2.nufone.net;tag=as4a099353

 Call-ID: 401507385 at 66.159.194.51

 CSeq: 3364 REGISTER

 User-Agent: Asterisk PBX

 Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER

 Contact: <sip:seberino at 198.22.67.70>

 Content-Length: 0




2005/01/22 10:32 PST [ui] got SIP response 100: Trying
2005/01/22 10:32 PST [ui] got SIP response
  SIP/2.0 100 Trying

 Via: SIP/2.0/UDP 66.159.194.51:5060;received=66.159.194.51;rport=5060

 From: sip:seberino at switch-2.nufone.net

 To: sip:seberino at switch-2.nufone.net;tag=as4a099353

 Call-ID: 401507385 at 66.159.194.51

 CSeq: 3364 REGISTER

 User-Agent: Asterisk PBX

 Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER

 Contact: <sip:seberino at 198.22.67.70>

 Content-Length: 0




2005/01/22 10:32 PST [ui] Got a SIP packet from 198.22.67.70:5060
2005/01/22 10:32 PST [ui] SIP PACKET
 SIP/2.0 200 OK

 Via: SIP/2.0/UDP 66.159.194.51:5060;received=66.159.194.51;rport=5060

 From: sip:seberino at switch-2.nufone.net

 To: sip:seberino at switch-2.nufone.net;tag=as4a099353

 Call-ID: 401507385 at 66.159.194.51

 CSeq: 3364 REGISTER

 User-Agent: Asterisk PBX

 Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER

 Expires: 900

 Contact: <sip:seberino at 66.159.194.51:5060>;expires=900

 Date: Sat, 22 Jan 2005 18:32:26 GMT

 Content-Length: 0




2005/01/22 10:32 PST [ui] got SIP response 200: OK
2005/01/22 10:32 PST [ui] got SIP response
  SIP/2.0 200 OK

 Via: SIP/2.0/UDP 66.159.194.51:5060;received=66.159.194.51;rport=5060

 From: sip:seberino at switch-2.nufone.net

 To: sip:seberino at switch-2.nufone.net;tag=as4a099353

 Call-ID: 401507385 at 66.159.194.51

 CSeq: 3364 REGISTER

 User-Agent: Asterisk PBX

 Allow: INVITE, ACK, CANCEL, OPTIONS, BYE, REFER

 Expires: 900

 Contact: <sip:seberino at 66.159.194.51:5060>;expires=900

 Date: Sat, 22 Jan 2005 18:32:26 GMT

 Content-Length: 0


From anthony at interlink.com.au  Sun Jan 23 06:18:53 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Sun, 23 Jan 2005 16:18:53 +1100
Subject: [Shtoom] Shtoom/SIP working (thru NAT!) ????
In-Reply-To: <20050122183937.GA10154@spawar.navy.mil>
References: <5E11CDD6-6C91-11D9-9811-000A95E2A184@zooko.com>
	<20050122183937.GA10154@spawar.navy.mil>
Message-ID: <200501231618.54347.anthony@interlink.com.au>

On Sunday 23 January 2005 05:39, seberino at spawar.navy.mil wrote:
> Is Shtoom working behind my NATing firewall??  Looks like perhaps yes...

There's been a lot of work on this recently - please svn up, and 
include the output of shtoominfo.py. 

Anthony

-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From zooko at zooko.com  Sun Jan 23 12:35:07 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 23 Jan 2005 07:35:07 -0400
Subject: [Shtoom] patch: refactor packets
Message-ID: <D4E461F4-6D32-11D9-B71B-000A95E2A184@zooko.com>

Anthony:

This patch (as discussed on IRC) refactors the RTP Packet object and 
the parser in a way that I like.

It separates the PT (Payload Type) byte from the Content Type object.  
It makes parsing a stateless function instead of a parser object.  It 
makes all values of an RTP packet passed as ctor args instead of poked 
into place after construction.  It handles up to one RTP extension 
header.  It puts the RTP header stuff into a separate object, which is 
convenient if you have a function that needs to get the netbytes for 
the header separately from the netbytes for the payload (which I have).

This patch also refactors the "send an RTP packet" functionality of RTP 
Protocol.  There is a nice unified "_send_packet()" and a nice 
"_send_cn_packet()".

I've been using this patch on my proprietary branch since last October. 
  Shtoom passes unit tests after application of this patch.  I made a 
test call with Shtoom after application of this patch and it worked.

Further patches will be forthcoming which depend on and complement this 
patch.

Please apply.

Regards,

Zooko

The patch is inlined *and* attached:

P.S.  If you don't like nested classes, move Header outside of Packet, 
or mention it and I'll do so.

Index: shtoom/test/test_playout.py
===================================================================
--- shtoom/test/test_playout.py	(revision 1157)
+++ shtoom/test/test_playout.py	(working copy)
@@ -68,8 +68,7 @@
                  ts = 0
                  seq = 10017
                  while True:
-                    p = RTPPacket(data='', pt=None, ts=ts)
-                    p.seq = seq
+                    p = RTPPacket(0, seq, ts, '')
                      yield p
                      ts += 160
                      seq += 1
Index: shtoom/test/test_codecs.py
===================================================================
--- shtoom/test/test_codecs.py	(revision 1157)
+++ shtoom/test/test_codecs.py	(working copy)
@@ -56,10 +56,10 @@
          ae = self.assertEquals
          ae(p.encode('frobozulate'), 'frobozulate')
          ae(p.decode('frobozulate'), 'frobozulate')
-        p = RTPPacket(PT_RAW, 'farnarkling', ts=None)
+        p = RTPPacket(0, 0, 0, 'farnarkling', ct=PT_RAW)
          ae(c.decode(p), 'farnarkling')
          ae(c.encode('farnarkling').data, 'farnarkling')
-        ae(c.encode('farnarkling').pt, PT_RAW)
+        ae(c.encode('farnarkling').header.pt, PT_RAW.pt)

      # XXX testing other codecs - endianness issues? crap.

@@ -72,7 +72,7 @@
          ae(c.getDefaultFormat(), PT_PCMU)
          ae(len(c.encode(instr).data), 160)
          ae(c.encode(instr).data, ulawout)
-        ae(c.encode(instr).pt, PT_PCMU)
+        ae(c.encode(instr).header.ct, PT_PCMU)

      def testGSMCodec(self):
          if codecs.gsm is None:
Index: shtoom/test/test_rtp.py
===================================================================
--- shtoom/test/test_rtp.py	(revision 1157)
+++ shtoom/test/test_rtp.py	(working copy)
@@ -45,33 +45,29 @@
              ae(pt, RTPDict[(pt.name.lower(),pt.clock,pt.params or 1)])

      def testRTPPacketRoundTrip(self):
-        from shtoom.rtp.packets import RTPParser, RTPPacket
+        from shtoom.rtp.packets import RTPPacket
+        from shtoom.rtp.protocol import parse_rtppacket
          from shtoom.rtp.formats import PT_PCMU, PT_SPEEX, PT_GSM, PT_CN
          ae = self.assertEquals

-        ptdict = {}
-        for pt, byte in ((PT_PCMU, 
0),(PT_GSM,3),(PT_SPEEX,101),(PT_CN,13)):
-            ptdict[pt] = byte
-            ptdict[byte] = pt
-        parser = RTPParser(ptdict)
          ts = 12345678
          seq = 12345
          ssrc = 100001
          packets = [
-            RTPPacket(PT_PCMU, ''.join([chr(x) for x in range(160)]), 
0),
-            RTPPacket(PT_GSM, '\0'*33, 0, 1),
-            RTPPacket(PT_GSM, '\0'*33, 0, 0),
-            RTPPacket(PT_CN, chr(127), 0, 0),
-            RTPPacket(PT_CN, chr(127), 0, 1),
+            RTPPacket(ssrc, seq, ts, ''.join([chr(x) for x in 
range(160)]), 0, ct=PT_PCMU),
+            RTPPacket(ssrc, seq, ts, '\0'*33, 1, ct=PT_GSM),
+            RTPPacket(ssrc, seq, ts, '\0'*33, 0, ct=PT_GSM),
+            RTPPacket(ssrc, seq, ts, chr(127), 0, ct=PT_CN),
+            RTPPacket(ssrc, seq, ts, chr(127), 1, ct=PT_CN),
                    ]
          for pack in packets:
-            rpack = parser.fromnet(parser.tonet(pack, seq, ts, ssrc), 
None)
-            ae(pack.pt, rpack.pt)
-            ae(pack.marker, rpack.marker)
+            rpack = parse_rtppacket(pack.netbytes())
+            ae(pack.header.pt, rpack.header.pt)
+            ae(pack.header.marker, rpack.header.marker)
              ae(pack.data, rpack.data)
-            ae(rpack.seq, seq)
-            ae(rpack.ts, ts)
-            ae(rpack.ssrc, ssrc)
+            ae(rpack.header.seq, seq)
+            ae(rpack.header.ts, ts)
+            ae(rpack.header.ssrc, ssrc)

      def testSDPGen(self):
          from shtoom.rtp.formats import SDPGenerator, PTMarker
Index: shtoom/stun.py
===================================================================
--- shtoom/stun.py	(revision 1157)
+++ shtoom/stun.py	(working copy)
@@ -16,7 +16,7 @@
  from shtoom.defcache import DeferredCache
  from shtoom.nat import BaseMapper

-STUNVERBOSE = True
+STUNVERBOSE = False
  # If we're going to follow RFC recommendation, make this 7
  MAX_RETRANSMIT = 5

@@ -397,7 +397,10 @@
          else:
              if STUNVERBOSE:
                  print "giving up on %r"%(address,)
-            del self._potentialStuns[tid]
+            if hasattr(self, '_potentialStuns') and 
self._potentialStuns.has_key(tid):
+                del self._potentialStuns[tid]
+            else:
+                print "xxxxx what's going on here?  getattr(self, 
'_potentialStuns'): %s, val: %s" % (getattr(self, '_potentialStuns'), 
getattr(self, '_potentialStuns', {}).get(tid),)
              if not self._potentialStuns:
                  if STUNVERBOSE:
                      print "stun state 1 timeout - no internet UDP 
possible"
Index: shtoom/app/phone.py
===================================================================
--- shtoom/app/phone.py	(revision 1157)
+++ shtoom/app/phone.py	(working copy)
@@ -35,7 +35,7 @@
          self._currentCall = None
          self._muted = False
          self._rtpProtocolClass = None
-        self._debugrev = 10
+        self._debugrev = 17

      def needsThreadedUI(self):
          return self.ui.threadedUI
Index: shtoom/audio/converters.py
===================================================================
--- shtoom/audio/converters.py	(revision 1157)
+++ shtoom/audio/converters.py	(working copy)
@@ -202,15 +202,15 @@
          encaudio = codec.encode(bytes)
          if not encaudio:
              return None
-        return RTPPacket(self.format, encaudio, ts=None)
+        return RTPPacket(0, 0, 0, encaudio, ct=self.format)

      def decode(self, packet):
          "Accepts an RTPPacket, emits audio as bytes"
          if not packet.data:
              return None
-        codec = self.codecs.get(packet.pt)
+        codec = self.codecs.get(packet.header.ct)
          if not codec:
-            raise ValueError("can't decode format %r"%packet.pt)
+            raise ValueError("can't decode format %r"%packet.header.ct)
          encaudio = codec.decode(packet.data)
          return encaudio

Index: shtoom/audio/playout.py
===================================================================
--- shtoom/audio/playout.py	(revision 1157)
+++ shtoom/audio/playout.py	(working copy)
@@ -61,22 +61,22 @@
              raise ValueError("playout got %s instead of 
bytes"%(type(bytes)))

          if self.expect_seq == None:
-            self.expect_seq = packet.seq # First packet. Initialize seq
+            self.expect_seq = packet.header.seq # First packet. 
Initialize seq

          backlog = len(self.queue)
-        if packet.seq == self.expect_seq:
-            self.expect_seq = packet.seq+1
+        if packet.header.seq == self.expect_seq:
+            self.expect_seq = packet.header.seq+1
              if backlog < self.backlog+1:
                  self.queue.append(bytes)
              elif DEBUG:
                  log.msg("BacklogPlayout discarding")
          else:
-            offset = packet.seq-self.expect_seq
+            offset = packet.header.seq-self.expect_seq
              if offset > 0 and offset < 3:
                  # Fill with empty packets
                  self.queue += [None]*offset
                  self.queue.append(bytes)
-                self.expect_seq = packet.seq+1
+                self.expect_seq = packet.header.seq+1
                  if DEBUG:
                      log.msg("BacklogPlayout got hole at %d"%offset)
              elif offset < 0 and offset > -backlog:
Index: shtoom/rtp/packets.py
===================================================================
--- shtoom/rtp/packets.py	(revision 1157)
+++ shtoom/rtp/packets.py	(working copy)
@@ -6,75 +6,80 @@
  import struct
  from time import sleep, time

+# This class supports extension headers, but only one per packet.
  class RTPPacket:
-    "An RTPPacket contains RTP data to/from the RTP object"
-    def __init__(self, pt, data, ts, marker=0):
-        self.pt = pt
-        self.data = data
-        self.ts = ts
-        self.marker = marker
-        self.ssrc = None
-        self.seq = None
+    """ Contains RTP data. """
+    class Header:
+        def __init__(self, ssrc, pt, ct, seq, ts, marker=0, 
xhdrtype=None, xhdrdata=''):
+            """
+            If xhdrtype is not None then it is required to be an int 
 >= 0 and < 2**16 and xhdrdata is required to be a string.
+            """
+            assert isinstance(ts, (int, long,)), "ts: %s :: %s" % (ts, 
type(ts),)
+            assert isinstance(ssrc, (int, long,))
+            assert xhdrtype is None or isinstance(xhdrtype, int) and 
xhdrtype >= 0 and xhdrtype < 2**16
+            assert xhdrtype is None or isinstance(xhdrdata, str)

-    def __repr__(self):
-        return "<RTPPacket containing %r at %x>"%(self.pt, id(self))
+            self.ssrc, self.pt, self.ct, self.seq, self.ts, 
self.marker, self.xhdrtype, self.xhdrdata = ssrc, pt, ct, seq, ts, 
marker, xhdrtype, xhdrdata

+        def netbytes(self):
+            "Return network-formatted header."
+            assert isinstance(self.pt, int) and self.pt >= 0 and 
self.pt < 2**8, "pt is required to be a simple byte, suitable for 
stuffing into an RTP packet and sending. pt: %s" % self.pt
+            if self.xhdrtype is not None:
+                firstbyte = 0x90
+                xhdrnetbytes = struct.pack('!HH', self.xhdrtype, 
len(self.xhdrdata)) + self.xhdrdata
+            else:
+                firstbyte = 0x80
+                xhdrnetbytes = ''
+            return struct.pack('!BBHII', firstbyte, self.pt | 
self.marker << 7, self.seq % 2**16, self.ts, self.ssrc) + xhdrnetbytes

-class RTPParser:
-    """ An RTPParser creates RTPPacket objects from a bytestring. It is
-        created with a mapping of RTP PT bytes to PT markers"""
-    def __init__(self, ptdict):
-        self.ptdict = ptdict
+    def __init__(self, ssrc, seq, ts, data, pt=None, ct=None, 
marker=0, authtag='', xhdrtype=None, xhdrdata=''):
+        assert pt is None or isinstance(pt, int) and pt >= 0 and pt < 
2**8, "pt is required to be a simple byte, suitable for stuffing into 
an RTP packet and sending. pt: %s" % pt
+        self.header = RTPPacket.Header(ssrc, pt, ct, seq, ts, marker, 
xhdrtype, xhdrdata)
+        self.data = data
+        self.authtag = authtag # please leave this alone even if it 
appears unused -- it is required for SRTP

-    def tonet(self, packet, seq, ts, ssrc):
-        "Return network formatted packet"
-        pt = packet.pt
-        # XXX handle cc and x!
-        byte0 = 0x80
-        fmt = self.ptdict[pt] & 127
-        if packet.marker: fmt = fmt | 128
-        hdr = struct.pack('!BBHII', byte0, fmt, seq, ts, ssrc)
-        # Padding?
-        return hdr + packet.data
+    def __repr__(self):
+        if self.header.ct is not None:
+            ptrepr = "%r" % (self.header.ct,)
+        else:
+            ptrepr = "pt %d" % (self.header.pt,)

-    # XXX Note that the MediaLayer will create it's own RTPPackets - 
this is
-    # purely for packets coming off the network.
-    def fromnet(self, bytes, fromaddr):
-        hdr = struct.unpack('!BBHII', bytes[:12])
-        padding = hdr[0]&32 and 1 or 0
-        cc = hdr[0]&15 and 1 or 0
-        x = hdr[0]&16 and 1 or 0
-        pt = hdr[1]&127
-        if not self.ptdict.get(pt):
-            # Could be any old garbage, such as a late STUN packet
-            return None
-        pt = self.ptdict[pt]
-        marker = hdr[1]&128 and 1 or 0
-        seq = hdr[2]
-        ts = hdr[3]
-        ssrc = hdr[4]
-        headerlen = 12
-        if cc > 1:
-            headerlen = headerlen + 4 * cc
-        data = bytes[headerlen:]
-        if x:
-            # Mmm. Tasty tasty header extensions. We eats them.
-            xhdrtype,xhdrlen = struct.unpack('!HH', data[:4])
-            data = data[4+4*xhdrlen:]
-        if padding:
-            padcount = ord(data[-1])
-            if padcount:
-                data = data[:-padcount]
-        packet = RTPPacket(pt, data, ts, marker)
-        packet.ssrc = ssrc
-        packet.seq = seq
-        return packet
+        if self.header.xhdrtype is not None:
+            return "<%s #%d (%s) %s [%s] at 
%x>"%(self.__class__.__name__, self.header.seq, self.header.xhdrtype, 
ptrepr, repr(self.header.xhdrdata), id(self))
+        else:
+            return "<%s #%d %s at %x>"%(self.__class__.__name__, 
self.header.seq, ptrepr, id(self))

-    def haspt(self, key):
-        return self.ptdict.has_key(key)
+    def netbytes(self):
+        "Return network-formatted packet."
+        return self.header.netbytes() + self.data + self.authtag

+def parse_rtppacket(bytes):
+    hdr = struct.unpack('!BBHII', bytes[:12])
+    padding = hdr[0]&32 and 1 or 0
+    cc = hdr[0]&15 and 1 or 0
+    x = hdr[0]&16 and 1 or 0
+    pt = hdr[1]&127
+    marker = hdr[1]&128 and 1 or 0
+    seq = hdr[2]
+    ts = hdr[3]
+    ssrc = hdr[4]
+    headerlen = 12
+    if cc > 1:
+        headerlen = headerlen + 4 * cc
+    data = bytes[headerlen:]
+    if x:
+        # Mmm. Tasty tasty header extensions. We eats them.  Except 
for the first one.
+        xhdrtype,xhdrlen = struct.unpack('!HH', data[:4])
+        xhdrdata = data[4:4+4*xhdrlen]
+        data = data[4+4*xhdrlen:]
+    else:
+        xhdrtype, xhdrdata = None, None
+    if padding:
+        padcount = ord(data[-1])
+        if padcount:
+            data = data[:-padcount]
+    return RTPPacket(ssrc, seq, ts, data, marker=marker, pt=pt, 
xhdrtype=xhdrtype, xhdrdata=xhdrdata)

-
  class NTE:
      "An object representing an RTP NTE (rfc2833)"
      # XXX at some point, this should be hooked into the 
RTPPacketFactory.
Index: shtoom/rtp/protocol.py
===================================================================
--- shtoom/rtp/protocol.py	(revision 1157)
+++ shtoom/rtp/protocol.py	(working copy)
@@ -12,7 +12,7 @@
  from twisted.python import log

  from shtoom.rtp.formats import SDPGenerator, PT_CN, PT_xCN
-from shtoom.rtp.packets import RTPPacket, RTPParser
+from shtoom.rtp.packets import RTPPacket, parse_rtppacket

  TWO_TO_THE_16TH = 2<<16
  TWO_TO_THE_32ND = 2<<32
@@ -46,14 +46,12 @@

      def setSDP(self, sdp):
          "This is the canonical SDP for the call"
-        from shtoom.rtp.packets import RTPParser
          self.app.selectDefaultFormat(self.cookie, sdp)
          rtpmap = sdp.getMediaDescription('audio').rtpmap
-        ptdict = {}
+        self.ptdict = {}
          for pt, (text, marker) in rtpmap.items():
-            ptdict[pt] = marker
-            ptdict[marker] = pt
-        self.rtpParser = RTPParser(ptdict)
+            self.ptdict[pt] = marker
+            self.ptdict[marker] = pt

      def createRTPSocket(self, locIP, needSTUN=False):
          """ Start listening on UDP ports for RTP and RTCP.
@@ -226,6 +224,26 @@
          d.addCallback(lambda x: self.rtpListener.stopListening())
          d.addCallback(lambda x: self.rtcpListener.stopListening())

+    def _send_packet(self, pt, data, xhdrtype=None, xhdrdata=''):
+        packet = RTPPacket(self.ssrc, self.seq, self.ts, data, pt=pt, 
xhdrtype=xhdrtype, xhdrdata=xhdrdata)
+
+        self.seq += 1
+        self.transport.write(packet.netbytes(), self.dest)
+
+    def _send_cn_packet(self):
+        # PT 13 is CN.
+        if self.ptdict.has_key(PT_CN):
+            cnpt = PT_CN.pt
+        elif self.ptdict.has_key(PT_xCN):
+            cnpt = PT_xCN.pt
+        else:
+            # We need to send SOMETHING!?!
+            cnpt = 0
+
+        log.msg("sending CN(%s) to seed firewall to %s:%d"%(cnpt, 
self.dest[0], self.dest[1]), system='rtp')
+
+        self._send_packet(cnpt, chr(127))
+
      def startSendingAndReceiving(self, dest, fp=None):
          self.dest = dest
          self.prevInTime = self.prevOutTime = time()
@@ -248,32 +266,14 @@
          self.LC.start(0.020)
          # Now send a single CN packet to seed any firewalls that might
          # need an outbound packet to let the inbound back.
-        # PT 13 is CN.
-        if self.rtpParser.haspt(PT_CN):
-            cnpt = PT_CN.pt
-        elif self.rtpParser.haspt(PT_xCN):
-            cnpt = PT_xCN.pt
-        else:
-            cnpt = None
-        if cnpt:
-            log.msg("sending CN(%s) to seed firewall to %s:%d"%(cnpt,
-                                    self.dest[0], self.dest[1]), 
system='rtp')
-            hdr = struct.pack('!BBHII', 0x80, cnpt, self.seq, 
self.ts,self.ssrc)
-            self.transport.write(hdr+chr(0), self.dest)
-        else:
-            log.msg("hack - seeding firewall with PT_PCMU")
-            hdr = struct.pack('!BBHII', 0x80, 0, self.seq, 
self.ts,self.ssrc)
-            self.transport.write(hdr+(160*chr(0)), self.dest)
-            # We need to send SOMETHING!?!
+        self._send_cn_packet()
          if hasattr(self.transport, 'connect'):
              self.transport.connect(*self.dest)

      def datagramReceived(self, datagram, addr):
          # XXX check for late STUN packets here. see, e.g 
datagramReceived in sip.py
-        if self.rtpParser is None:
-            log.msg("early(?) rtp packet, no rtpParser available")
-            return
-        packet = self.rtpParser.fromnet(datagram, addr)
+        packet = parse_rtppacket(datagram)
+        packet.header.ct = self.ptdict[packet.header.pt]
          if packet:
              self.app.receiveRTP(self.cookie, packet)

@@ -340,26 +340,19 @@
          # when we go from silent->talking, set the marker bit. Other 
end
          # can use this as an excuse to adjust playout buffer.
          if self.sample is not None:
-            packet = self.sample
+            pt = self.ptdict[self.sample.header.ct]
+            self._send_packet(pt, self.sample.data)
              self.sent += 1
-            data = self.rtpParser.tonet(packet, self.seq, self.ts, 
self.ssrc)
-            self.transport.write(data, self.dest)
              self.sample = None
-        else:
-            if (self.packets - self.sent) % 100 == 0:
-                if self.rtpParser.haspt(PT_CN):
-                    cn = RTPPacket(PT_CN, chr(127), 0)
-                    cn = self.rtpParser.tonet(cn, self.seq, self.ts, 
self.ssrc)
-                    self.transport.write(cn, self.dest)
-        self.seq += 1
+        elif (self.packets - self.sent) % 100 == 0:
+            self._send_cn_packet()
+
          # Now send any pending DTMF keystrokes
          if self._pendingDTMF:
              payload = self._pendingDTMF[0].getPayload(self.ts)
              if payload:
                  # XXX Hack. telephone-event isn't always 101!
-                hdr = struct.pack('!BBHII', 0x80, 101, self.seq, 
self.ts, self.ssrc)
-                self.transport.write(hdr+payload, self.dest)
-                self.seq += 1
+                self._send_packet(pt=101, data=payload)
                  if self._pendingDTMF[0].isDone():
                      self._pendingDTMF = self._pendingDTMF[1:]
          try:

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: refactor-packets.patch.txt
URL: <http://mail.python.org/pipermail/shtoom/attachments/20050123/99c74a5f/attachment.txt>

From anthony at interlink.com.au  Mon Jan 24 12:27:51 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Mon, 24 Jan 2005 22:27:51 +1100
Subject: [Shtoom] patch: refactor packets
In-Reply-To: <D4E461F4-6D32-11D9-B71B-000A95E2A184@zooko.com>
References: <D4E461F4-6D32-11D9-B71B-000A95E2A184@zooko.com>
Message-ID: <200501242227.52176.anthony@interlink.com.au>

On Sunday 23 January 2005 22:35, Zooko O'Whielacronx wrote:
> It separates the PT (Payload Type) byte from the Content Type object. ?
> It makes parsing a stateless function instead of a parser object. ?It
> makes all values of an RTP packet passed as ctor args instead of poked
> into place after construction. ?It handles up to one RTP extension
> header. ?It puts the RTP header stuff into a separate object, which is
> convenient if you have a function that needs to get the netbytes for
> the header separately from the netbytes for the payload (which I have).
>
> This patch also refactors the "send an RTP packet" functionality of RTP
> Protocol. ?There is a nice unified "_send_packet()" and a nice
> "_send_cn_packet()".

This is now checked in. Thanks for the patch!

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From zooko at zooko.com  Tue Jan 25 19:26:32 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 25 Jan 2005 14:26:32 -0400
Subject: [Shtoom] playout algorithm
Message-ID: <A33D4230-6EFE-11D9-9370-000A95E2A184@zooko.com>

I noticed a comment about anthony wanting to improve the playout 
object.  I've written a completely new playout object which I've used 
successfully for months now, but only with PCMU.  I'll be submitting it 
in a patch via e-mail very soon.

--Z



From zooko at zooko.com  Tue Jan 25 20:58:00 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 25 Jan 2005 15:58:00 -0400
Subject: [Shtoom] local ip determination
Message-ID: <6A90B42E-6F0B-11D9-9370-000A95E2A184@zooko.com>

Should sip.Call.setLocalIP() be refactored to use the new 
nat.getLocalIPAddress()?

By the way, you might be able to glean useful ideas or code from this 
venerable and battle-tested module:

http://mnetproject.org/repos/mnet/egtp/ipaddresslib.py

I don't really know most of what is in that hairy module, but I do know 
that it worked well enough on all supported platforms at Mojo Nation 
and then Mnet for many years.

The authors Greg Smith and Hauke Johannknecht (with help from others) 
have my respect for their network fu.

--Z



From anthony at interlink.com.au  Wed Jan 26 10:56:49 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 26 Jan 2005 20:56:49 +1100
Subject: [Shtoom] local ip determination
In-Reply-To: <6A90B42E-6F0B-11D9-9370-000A95E2A184@zooko.com>
References: <6A90B42E-6F0B-11D9-9370-000A95E2A184@zooko.com>
Message-ID: <200501262056.49814.anthony@interlink.com.au>

On Wednesday 26 January 2005 06:58, Zooko O'Whielacronx wrote:
> Should sip.Call.setLocalIP() be refactored to use the new
> nat.getLocalIPAddress()?
>
> By the way, you might be able to glean useful ideas or code from this
> venerable and battle-tested module:
>
> http://mnetproject.org/repos/mnet/egtp/ipaddresslib.py
>
> I don't really know most of what is in that hairy module, but I do know
> that it worked well enough on all supported platforms at Mojo Nation
> and then Mnet for many years.

Hm. They try the connected UDP socket, and then attempt to parse
ifconfig and netstat output. I really want to stay away from the latter -
it strikes me as an incredibly brittle and painful approach (indeed, the
module has a huge pile of parsing code for all sorts of different platforms
to parse the output. Yuk.)

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From anthony at interlink.com.au  Wed Jan 26 10:58:40 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 26 Jan 2005 20:58:40 +1100
Subject: [Shtoom] playout algorithm
In-Reply-To: <A33D4230-6EFE-11D9-9370-000A95E2A184@zooko.com>
References: <A33D4230-6EFE-11D9-9370-000A95E2A184@zooko.com>
Message-ID: <200501262058.40902.anthony@interlink.com.au>

On Wednesday 26 January 2005 05:26, Zooko O'Whielacronx wrote:
> I noticed a comment about anthony wanting to improve the playout
> object.  I've written a completely new playout object which I've used
> successfully for months now, but only with PCMU.  I'll be submitting it
> in a patch via e-mail very soon.

Great! For now, I've switched the default back to BrainDeadPlayout - 
I've received a number of complaints about the BacklogPlayout causing 
audio to just stop after several minutes of talking, and I haven't the time 
to dig into why it's failing.

-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.


From zooko at zooko.com  Sun Jan 30 21:40:16 2005
From: zooko at zooko.com (Zooko O'Whielacronx)
Date: 30 Jan 2005 16:40:16 -0400
Subject: [Shtoom] upnp leaking mappings
Message-ID: <aebd4a37b94ba5418b97052ce28d4a50@zooko.com>

Whenever I do UPnP, I see a list of "existing mappings" which is one 
greater than the last time I did UPnP.

Also the most recent time I did it my linksys didn't respond to the 
upnp discovery broadcast at all until I disabled and re-enabled upnp in 
the linksys admin web page.

I had assumed that these mappings would time-out if they weren't 
refreshed, but they've lived for days now.

Ought they to be cleaned up, and if so how?

--Z



From tedbukov at gmail.com  Sun Jan 30 23:36:42 2005
From: tedbukov at gmail.com (Todor Bukov)
Date: Sun, 30 Jan 2005 22:36:42 +0000
Subject: [Shtoom] About the present shtoom/doug status
Message-ID: <3e1091d050130143656aaf85d@mail.gmail.com>

Hi folks,

I was looking for SIP/voice libraries to include it into a Python
project and I was quite surprised to find out that there is already a
complete application  - shtoom. I read the paper in python.org and the
project home page, but I found that 0.2 version is rather old and 0.3
is yet to come. Could someone give me more information about the
current status of the project and when the next version is expeted to
be released? I also consider to get involved into this project and
would like to review the documentation. I'll be very thankfull if you
inform me whether the project is still in active development and who
maintain it (is sourceforge.net information up to date)?

 Thanks in advance for your answer.
--- Ted ---


From anthony at interlink.com.au  Mon Jan 31 13:58:48 2005
From: anthony at interlink.com.au (Anthony Baxter)
Date: Mon, 31 Jan 2005 23:58:48 +1100
Subject: [Shtoom] About the present shtoom/doug status
In-Reply-To: <3e1091d050130143656aaf85d@mail.gmail.com>
References: <3e1091d050130143656aaf85d@mail.gmail.com>
Message-ID: <200501312358.49620.anthony@interlink.com.au>

On Monday 31 January 2005 09:36, Todor Bukov wrote:
> I was looking for SIP/voice libraries to include it into a Python
> project and I was quite surprised to find out that there is already a
> complete application  - shtoom. I read the paper in python.org and the
> project home page, but I found that 0.2 version is rather old and 0.3
> is yet to come.

Yes - getting a new release out has turned into quite the nightmare.
I'm working very hard on finishing off the bits I still have to do to 
make this happen. 

> Could someone give me more information about the 
> current status of the project and when the next version is expeted to
> be released? 

"Soon" is the best answer I can give you, I'm afraid - I'd be very
disappointed if it's not during February some time. Having said 
that, the current SVN trunk (see instructions on the webpage at
shtoom.divmod.org) is pretty stable, and there's been a huge amount
of stuff added since 0.2

> I also consider to get involved into this project and 
> would like to review the documentation. I'll be very thankfull if you
> inform me whether the project is still in active development and who
> maintain it (is sourceforge.net information up to date)?

The project is no longer at sourceforge, but at shtoom.divmod.org.
It's very much a live project.

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>
It's never too late to have a happy childhood.