Document request - ROSE protocol specification

Document request - ROSE protocol specification

[Richard Stearn]

Greeting all

I am attempting to add ROSE to wireshark (ethereal).  My preference is
to code the protocol analyser from the specification document rather
than reverse engineer an existing driver stack.

The RATS web site www.rats.org now goes to a holding site.  Does anyone
have either an electronic copy of the ROSE spec or a URL to a current
copy?

TIA
-- 
Regards
	Richard

Re: Document request - ROSE protocol specification

[Folkert van Heusden]

Hi,

> Greeting all
> I am attempting to add ROSE to wireshark (ethereal).  My preference is
> to code the protocol analyser from the specification document rather
> than reverse engineer an existing driver stack.
> The RATS web site www.rats.org now goes to a holding site.  Does anyone
> have either an electronic copy of the ROSE spec or a URL to a current
> copy?

Would you please also consider integrating all this work into tcpdump?
Because as far as I know wireshark is for x11.


Folkert van Heusden

MultiTail er et flexible tool for å kontrolere Logfiles og commandoer.
Med filtrer, farger, sammenføringer, forskeliger ansikter etc.
http://www.vanheusden.com/multitail/
----------------------------------------------------------------------
Phone: +31-6-41278122, PGP-key: 1F28D8AE, www.vanheusden.com
-
To unsubscribe from this list: send the line "unsubscribe linux-hams" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Document request - ROSE protocol specification

[Richard Stearn]

Folkert van Heusden wrote:

> Would you please also consider integrating all this work into tcpdump?
> Because as far as I know wireshark is for x11.

You can build and run tshark which is the terminal version of wireshark.

I have a patch for tcpdump that contains the changes to support:
	AX.25
	AX.25 ARP
	Netrom
	Flexnet (such as it is)
	ROSE (such as it is)

I will release that at the same time as the wireshark one.  The full set of 
patches is:
	libpcap - for protocol 3 (AX.25) capture
	tcpdump
	wireshark

I originally started by adding to tcpdump as it is a simpler system and I could 
then use the experience gained when porting to wireshark.

-- 
Regards
	Richard

Re: Document request - ROSE protocol specification

[Chuck Hast]

On 3/14/07, Richard Stearn <richard@rns-stearn.demon.co.uk> wrote:
> Greeting all
>
> I am attempting to add ROSE to wireshark (ethereal).  My preference is
> to code the protocol analyser from the specification document rather
> than reverse engineer an existing driver stack.
>
> The RATS web site www.rats.org now goes to a holding site.  Does anyone
> have either an electronic copy of the ROSE spec or a URL to a current
> copy?
>

Richard,
I googled on X.25 PLP and found that most of the descriptors pretty much
follow what I recall from Brian Lantz decoding of ROSE in TNOS. Listen
does not show the GFI just the LCN. Here is a good website for the X.25
PLP breakout:

http://www.techfest.com/networking/wan/x25plp.htm

Also perhaps Bernard F6BVP can provide you information or Jean-Paul
F6FBB, Bernard is actively working on FPAC (the set of tools and wrapper
that makes ROSE easy to use under Linux and also allows for integration
of IP and NetRom under a single system).

Back to X.25 PLP, there is a lot of info out there and again as far as I
can remember, the only things that may not have been always visible
were all of the different cause diagnostic codes because there are a
lot of them that deal with commercial operations.

-- 
Chuck Hast  -- KP4DJT --
To paraphrase my flight instructor;
"the only dumb question is the one you DID NOT ask resulting in my going
out and having to identify your bits and pieces in the midst of torn
and twisted metal."

Re: Document request - ROSE protocol specification

[Bernard Pidoux]

Chuck Hast wrote:
> On 3/14/07, Richard Stearn <richard@rns-stearn.demon.co.uk> wrote:
>> Greeting all
>>
>> I am attempting to add ROSE to wireshark (ethereal).  My preference is
>> to code the protocol analyser from the specification document rather
>> than reverse engineer an existing driver stack.
>>
>> The RATS web site www.rats.org now goes to a holding site.  Does anyone
>> have either an electronic copy of the ROSE spec or a URL to a current
>> copy?
>>
> 
> Richard,
> I googled on X.25 PLP and found that most of the descriptors pretty much
> follow what I recall from Brian Lantz decoding of ROSE in TNOS. Listen
> does not show the GFI just the LCN. Here is a good website for the X.25
> PLP breakout:
> 
> http://www.techfest.com/networking/wan/x25plp.htm
> 
> Also perhaps Bernard F6BVP can provide you information or Jean-Paul
> F6FBB, Bernard is actively working on FPAC (the set of tools and wrapper
> that makes ROSE easy to use under Linux and also allows for integration
> of IP and NetRom under a single system).
> 
> Back to X.25 PLP, there is a lot of info out there and again as far as I
> can remember, the only things that may not have been always visible
> were all of the different cause diagnostic codes because there are a
> lot of them that deal with commercial operations.
> 

ROSE has been described by Tom Moulton W2VY in different ARRL Computer
Networking Conference proceedings available on CDs (google ROSE W2VY).
Sorry I only have 3rd (1984), 9th (1990) and 13th (1994) books
                                                     without ROSE
protocole descripion.








ROSE adressing follows the X.121 specs and uses 14 digits.

There is a good description here :
http://www.rhyshaden.com/x25.htm

Linux kernel source provides diagnostic codes and specs.
For example in /linux/include/net/rose.h :


#define ROSE_ADDR_LEN                   5

#define ROSE_MIN_LEN                    3

#define ROSE_GFI                        0x10
#define ROSE_Q_BIT                      0x80
#define ROSE_D_BIT                      0x40
#define ROSE_M_BIT                      0x10

#define ROSE_CALL_REQUEST               0x0B
#define ROSE_CALL_ACCEPTED              0x0F
#define ROSE_CLEAR_REQUEST              0x13
#define ROSE_CLEAR_CONFIRMATION         0x17
#define ROSE_DATA                       0x00
#define ROSE_INTERRUPT                  0x23
#define ROSE_INTERRUPT_CONFIRMATION     0x27
#define ROSE_RR                         0x01
#define ROSE_RNR                        0x05
#define ROSE_REJ                        0x09
#define ROSE_RESET_REQUEST              0x1B
#define ROSE_RESET_CONFIRMATION         0x1F
#define ROSE_REGISTRATION_REQUEST       0xF3
#define ROSE_REGISTRATION_CONFIRMATION  0xF7
#define ROSE_RESTART_REQUEST            0xFB
#define ROSE_RESTART_CONFIRMATION       0xFF
#define ROSE_DIAGNOSTIC                 0xF1
#define ROSE_ILLEGAL                    0xFD


ROSE packets identified by listen have PID=1(X.25)

Here is a sample screen dump from F6BVP FPAC node F6BVP-11 :

0: fm KP4DJT-9 to F6BVP-11 ctl I27^ pid=1(X.25) len 3
X.25: LCI 032 : RR R7
0: fm KP4DJT-9 to F6BVP-11 ctl RR3+
0: fm F1MVP-5 to F6BVP-11 ctl SABM+
0: fm F1MVP-5 to F6BVP-11 ctl I00+ pid=1(X.25) len 5
X.25: LCI 000 : RESTART REQUEST - Cause DTE Originated - Diag 0
0: fm VK2TV-2 to F6BVP-11 ctl RR0v
0: fm F5KBW-9 to F6BVP-11 ctl SABM+
0: fm F5KBW-9 to F6BVP-11 ctl I00^ pid=1(X.25) len 5
X.25: LCI 000 : RESTART REQUEST - Cause DTE Originated - Diag 0
0: fm F5KBW-9 to F6BVP-11 ctl I11^ pid=1(X.25) len 60
X.25: LCI 064 : CALL REQUEST - NbAlea: B464
fm F5KBW-3   @2080,833501
to F6BVP-3   @2080,175502


73 de Bernard, f6bvp

Re: Document request - ROSE protocol specification

[Chuck Hast]

On 3/15/07, Bernard Pidoux <pidoux@ccr.jussieu.fr> wrote:
>
> Chuck Hast wrote:
> > On 3/14/07, Richard Stearn <richard@rns-stearn.demon.co.uk> wrote:
> >> Greeting all
> >>
> >> I am attempting to add ROSE to wireshark (ethereal).  My preference is
> >> to code the protocol analyser from the specification document rather
> >> than reverse engineer an existing driver stack.
> >>
> >> The RATS web site www.rats.org now goes to a holding site.  Does anyone
> >> have either an electronic copy of the ROSE spec or a URL to a current
> >> copy?
> >>
> >
> > Richard,
> > I googled on X.25 PLP and found that most of the descriptors pretty much
> > follow what I recall from Brian Lantz decoding of ROSE in TNOS. Listen
> > does not show the GFI just the LCN. Here is a good website for the X.25
> > PLP breakout:
> >
> > http://www.techfest.com/networking/wan/x25plp.htm
> >
> > Also perhaps Bernard F6BVP can provide you information or Jean-Paul
> > F6FBB, Bernard is actively working on FPAC (the set of tools and wrapper
> > that makes ROSE easy to use under Linux and also allows for integration
> > of IP and NetRom under a single system).
> >
> > Back to X.25 PLP, there is a lot of info out there and again as far as I
> > can remember, the only things that may not have been always visible
> > were all of the different cause diagnostic codes because there are a
> > lot of them that deal with commercial operations.
> >
>
> ROSE has been described by Tom Moulton W2VY in different ARRL Computer
> Networking Conference proceedings available on CDs (google ROSE W2VY).
> Sorry I only have 3rd (1984), 9th (1990) and 13th (1994) books
>                                                      without ROSE
> protocole descripion.
>
>
>
>
>
>
>
>
> ROSE adressing follows the X.121 specs and uses 14 digits.
>
> There is a good description here :
> http://www.rhyshaden.com/x25.htm
>
> Linux kernel source provides diagnostic codes and specs.
> For example in /linux/include/net/rose.h :
>
>
> #define ROSE_ADDR_LEN                   5
>
> #define ROSE_MIN_LEN                    3
>
> #define ROSE_GFI                        0x10
> #define ROSE_Q_BIT                      0x80
> #define ROSE_D_BIT                      0x40
> #define ROSE_M_BIT                      0x10
>
> #define ROSE_CALL_REQUEST               0x0B
> #define ROSE_CALL_ACCEPTED              0x0F
> #define ROSE_CLEAR_REQUEST              0x13
> #define ROSE_CLEAR_CONFIRMATION         0x17
> #define ROSE_DATA                       0x00
> #define ROSE_INTERRUPT                  0x23
> #define ROSE_INTERRUPT_CONFIRMATION     0x27
> #define ROSE_RR                         0x01
> #define ROSE_RNR                        0x05
> #define ROSE_REJ                        0x09
> #define ROSE_RESET_REQUEST              0x1B
> #define ROSE_RESET_CONFIRMATION         0x1F
> #define ROSE_REGISTRATION_REQUEST       0xF3
> #define ROSE_REGISTRATION_CONFIRMATION  0xF7
> #define ROSE_RESTART_REQUEST            0xFB
> #define ROSE_RESTART_CONFIRMATION       0xFF
> #define ROSE_DIAGNOSTIC                 0xF1
> #define ROSE_ILLEGAL                    0xFD
>
>
> ROSE packets identified by listen have PID=1(X.25)
>
> Here is a sample screen dump from F6BVP FPAC node F6BVP-11 :
>
> 0: fm KP4DJT-9 to F6BVP-11 ctl I27^ pid=1(X.25) len 3
> X.25: LCI 032 : RR R7
> 0: fm KP4DJT-9 to F6BVP-11 ctl RR3+
> 0: fm F1MVP-5 to F6BVP-11 ctl SABM+
> 0: fm F1MVP-5 to F6BVP-11 ctl I00+ pid=1(X.25) len 5
> X.25: LCI 000 : RESTART REQUEST - Cause DTE Originated - Diag 0
> 0: fm VK2TV-2 to F6BVP-11 ctl RR0v
> 0: fm F5KBW-9 to F6BVP-11 ctl SABM+
> 0: fm F5KBW-9 to F6BVP-11 ctl I00^ pid=1(X.25) len 5
> X.25: LCI 000 : RESTART REQUEST - Cause DTE Originated - Diag 0
> 0: fm F5KBW-9 to F6BVP-11 ctl I11^ pid=1(X.25) len 60
> X.25: LCI 064 : CALL REQUEST - NbAlea: B464
> fm F5KBW-3   @2080,833501
> to F6BVP-3   @2080,175502
>
See, I told you that Bernard would have some good stuff... What listen does
not show is the GFI, it does show the LCI and in the call request packet the
random number  (NbAlea) which is used to assign the LCI. Nor does it show
the cause/diag code data when there is a error or a disconnect.

So any work you do in this direction will be a big help for us.


-- 
Chuck Hast  -- KP4DJT --
To paraphrase my flight instructor;
"the only dumb question is the one you DID NOT ask resulting in my going
out and having to identify your bits and pieces in the midst of torn
and twisted metal."

Re: Document request - ROSE protocol specification

[Richard Stearn]

Bernard Pidoux wrote:

> ROSE has been described by Tom Moulton W2VY in different ARRL Computer
> Networking Conference proceedings available on CDs (google ROSE W2VY).
> Sorry I only have 3rd (1984), 9th (1990) and 13th (1994) books
>                                                      without ROSE
> protocole descripion.
> 
> ROSE adressing follows the X.121 specs and uses 14 digits.
> 
> There is a good description here :
> http://www.rhyshaden.com/x25.htm
> 
> Linux kernel source provides diagnostic codes and specs.
> For example in /linux/include/net/rose.h :

Bernard

Thank you for your reply.  I think I am going have to follow the X.25
protocol stack upwards and find out what was changed the hard way.  The
good bit is that there is code in wireshark already (in the X.25 dissector)
that implements the PLP, so I will start from there.

-- 
Regards
	Richard

Re: Document request - ROSE protocol specification

[Richard Stearn]

Chuck Hast wrote:
> See, I told you that Bernard would have some good stuff... What listen does
> not show is the GFI, it does show the LCI and in the call request packet 
> the
> random number  (NbAlea) which is used to assign the LCI. Nor does it show
> the cause/diag code data when there is a error or a disconnect.
> 
> So any work you do in this direction will be a big help for us.

Chuck

There is an X.25 dissector in wireshark, I am looking at either calling or 
cloning the relevant parts.  The good news from your perspective is that it 
appears to be a comprehensive dissection.

-- 
Regards
	Richard

Re: Document request - ROSE protocol specification

[Chuck Hast]

On 3/16/07, Richard Stearn <richard@rns-stearn.demon.co.uk> wrote:
> Chuck Hast wrote:
> > See, I told you that Bernard would have some good stuff... What listen does
> > not show is the GFI, it does show the LCI and in the call request packet
> > the
> > random number  (NbAlea) which is used to assign the LCI. Nor does it show
> > the cause/diag code data when there is a error or a disconnect.
> >
> > So any work you do in this direction will be a big help for us.
>
> Chuck
>
> There is an X.25 dissector in wireshark, I am looking at either calling or
> cloning the relevant parts.  The good news from your perspective is that it
> appears to be a comprehensive dissection.
>
That is good, Tom Moulton who did the original ROSE code wrote X.25
PLP commercially, so he knew well the whole stack. Indeed he used to
write code for X.25 pad devices, and was a stickler for staying as near to
the protocol as possible. I would suspect that you will find that except for
the extensions to handle callsigns there is not that much difference but
for the commercial pieces which are not used.

If you want to set up a FPAC site there to test with, we can also help you
with that, particularly if you use axip to link to the rest of the network. Then
you will have a real live one to monitor.

-- 
Chuck Hast  -- KP4DJT --
To paraphrase my flight instructor;
"the only dumb question is the one you DID NOT ask resulting in my going
out and having to identify your bits and pieces in the midst of torn
and twisted metal."

Re: Document request - ROSE protocol specification

[Bernard Pidoux]

Chuck Hast wrote:

>...What listen does
> not show is the GFI, it does show the LCI and in the call request packet
> the
> random number  (NbAlea) which is used to assign the LCI. Nor does it show
> the cause/diag code data when there is a error or a disconnect.
> 
> So any work you do in this direction will be a big help for us.
> 
> 

Actually, listen provides some error diagnostic including the point
where the link failed as shown here :

0: fm KP4DJT-9 to F6BVP-11 ctl I72^ pid=1(X.25) len 26
X.25: LCI 031 : CLEAR REQUEST - Cause DTE Originated - Diag 0
at N3PPC-9 @3100,386335
0: fm KP4DJT-9 to F6BVP-11 ctl RR0v

Not so bad.

73 de Bernard, f6bvp

Re: Document request - ROSE protocol specification

[Chuck Hast]

On 3/18/07, Bernard Pidoux <pidoux@ccr.jussieu.fr> wrote:
>
>
> Chuck Hast wrote:
>
> >...What listen does
> > not show is the GFI, it does show the LCI and in the call request packet
> > the
> > random number  (NbAlea) which is used to assign the LCI. Nor does it show
> > the cause/diag code data when there is a error or a disconnect.
> >
> > So any work you do in this direction will be a big help for us.
> >
> >
>
> Actually, listen provides some error diagnostic including the point
> where the link failed as shown here :
>
> 0: fm KP4DJT-9 to F6BVP-11 ctl I72^ pid=1(X.25) len 26
> X.25: LCI 031 : CLEAR REQUEST - Cause DTE Originated - Diag 0
> at N3PPC-9 @3100,386335
> 0: fm KP4DJT-9 to F6BVP-11 ctl RR0v
>
> Not so bad.

Yes, that is how I knew that we were getting the error location information,
I was watching it in Listen and observed that on the fpac console it did
not match.  But being able to have a full set of tools in WireShark (was
Etherreal) for data collection and analysis would be a very good thing,
so I encourage him to get as much info as possible to integrate X.25 PLP
so we have one more tool to use.


-- 
Chuck Hast  -- KP4DJT --
To paraphrase my flight instructor;
"the only dumb question is the one you DID NOT ask resulting in my going
out and having to identify your bits and pieces in the midst of torn
and twisted metal."

Re: Document request - ROSE protocol specification

[Ralf Baechle DL5RB]

On Thu, Mar 15, 2007 at 12:22:10PM +0100, Bernard Pidoux wrote:

> > Also perhaps Bernard F6BVP can provide you information or Jean-Paul
> 
> ROSE adressing follows the X.121 specs and uses 14 digits.
> 
> There is a good description here :
> http://www.rhyshaden.com/x25.htm
> 
> Linux kernel source provides diagnostic codes and specs.
> For example in /linux/include/net/rose.h :

I would warn to take the Linux kernel code as a proper definition of the
ROSE protocol.  While Linux may be the dominant ROSE implementation today
the protocol code contains quite a few oddities that make me scratch the
head to say the least.

   Ralf