The fundamental problem with the AT code is there is no queuing system
implemented either to process messages taken off of the SIM or to send
messages out through the SIM, and the current 'try for a
while then die'
mechanism is far from perfect.
I think its a bit more than that. Here are
- logging facility: access.log data may be stored to mysql db,
especialy for automatic billing support.
- configuration facility: multi-groups like sms-user, sms-service,
ota-config may be stored in mysql db, so there may be added or
manipulated while runtime and the service would not have
Note the absence of WAP. If anybody actually cares about WAP, the WAP
components should be taken out and recast as a discrete product. The
original premise that WAP and SMS sit within a common stack
is not likely to
happen in the real world, and I sense that the two differing
SMSC's shouldn't interpret DCS field (That's why I'm pissed
with Siemens
M20)
It makes sense to me that GSM modems should interpret the DCS
because they have no responsibility for encoding the data.
GSM 03.38 chapter 4 makes it reasonably clear about when the DCS
value specifies 8 bit or
If other smsc's use standard dcs as of ETSI 03.38, they could just
use dcs = fields_to_dcs(msg, 0); and fields_to_dcs(dcs, msg);
when receiving a message.
I've patched all other smsc_*.c code (I hope I didn't introduce bugs )
so they already parse coding = DC_8BIT and coding = DC_UCS2
I did use several fields instead of a direct DCS value
because some smsc's
dont use it directly (I recall one sms'c that besides 7 and 8
bits, accepted
text in IA5 or something like that)
As far as I can see, only one SMSC (OIS) can't use it directly.
All the others
No it won't. :(
There's coding to define the data coding, 1=7bit, 2=8bit, 3=UCS2,
defaults to 1 or to 2 if udh is present.
so, just use udh=footext=barcoding=1
I've just being testing with some prototype EMS clients that required
DCS values of 0x11 and 0x15. Exposing DCS at the
I think you need to call
bb_smscconn_killed();
in the shutdown callback for the new AT driver. Otherwise
it looks like the sms_router() thread won't exit and the
shutdown hangs. Other drivers (SMPP, EMI2) make this
call in their shutdown callbacks.
Paul Keogh
ANAM Wireless Internet
You also need to call
bb_smscconn_sent ()
after a successful message send to ensure that the store is synch'd
(at least).
Andreas - I have a short word document that pulls all the SMSCConn
comments from the code together - are you interested in seeing it?
Paul Keogh
ANAM Wireless Internet
the code should be amended to OR 0x04 in for 8 bit only.
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle Yard, Saint Patrick's Road, Dalkey, County Dublin, Ireland
Is there anyway to bind to an ip-adress aswell as port?
I have severel ip-adresses, and I only want kannel to bind to one of
those, is this planned or already existent feature?
Thats what the wdp-interface-name variable is for in the BB config.
Its value is an IP address that the BB
i say this statment in the updated manual Siemens M20 (this
modem have
some bugs)
now this is bad for me seeing that i just order one, does
anyone know what
the bugs are?
Just make sure you get the latest version of the firmware (5.4)
and you'll minimise the bugs :-)
Our Kannel WAP gateway is exhibiting some strange behaviour
when visiting
a few sites. (Notably wap.ft.com).
The request gets sent off fine, and Kannel sees an HTTP 302 status
message, however it never seems to get the headers or the
Location header
and eventually the WAP device just
I've just noticed
else if (privdata-sendtype[i] == 31) {
warning(0, smsc_emi2: Alert (operation 31) was not
ACKed within %d seconds, privdata-waitack);
if (privdata-flowcontrol) write=1;
Then I've compiled with native malloc and I
got the same error.
You only see this with the gw_check_ interface and
you won't
getthis withnative malloc, so check your
build again.
Then I've added a throughput delay in EMI2 inside
the receive messages while, to
I'm now looking at your code.
You are forcing receiving sms's through ME memory, right ?
What about if some mobile doesn't work correctly with it ?
quote(SMSs go through memory - this is reliable, but only
implemented and
+ tested on an Ericsson gm25 modem and falcom */)
can should
I am working on a project to develop a GMSC for short
message service. I
have had a general study on the SS7 and the GSM-MAP. Now I
should select
a solution for communicating with the MSC and the HLR (note that the
physical layer is E1). I don't want to get involved in the
SS7
Started building support for client provisioning. Do you
know what Ericsson phones support WAP bookmarks ?
R320, R380 and R520 as far as I recall.
Do you mean WAP 2.0 client provisioning or Ericsson proprietary ?
Mainly WAP 2.0, but Ericsson proprietary would bo of some
Would the current CVS version of Kannel work with an SMSC
that only supports
SMPP v3.3?
Yes. Most (if not all) of the interworking I've done is with
3.3 SMSCs.
Hi. We've got access to an SMPP interface, which sort of works...
But the SMPP only allows us to do BIND_RECEIVER, not BIND_TRANSMITTER.
Is there any way to tell kannel to forget about that BIND_TRANSMIT
that it cannot have?
Kannel insists on doing an endless 2-second loop of
2002-01-09
You probably have to modify the RADIUS server to log the IP-CLI
mapping somewhere, and then have Kannel query this. Or it could
query the accounting logs.
I did some investigation of the various options for hooking CLI in a
RADIUS environment and concluded that the RADIUS protocol
Are there going to be any of us at 3GSM World Congress, 19-22.
February, Cannes, France?
Stipe
We (ANAM) will be there on the Enterprise Ireland stand. Drop
by and say hello !
Does anyone know if the the large data transfer extensions
are mandatory for MMS (or does the lack of them just restrict
the max. size of your messages ?). Also, any pointers to the
large data extensions specs ? The LDT WG was folded into WAP
NG at some point but I can't track it any further!
I read a conformance document about MMS. It does not mention
LDT but it does mention WTP SAR. Minimum message size is 30 kb.
The 2.0 spec. references large data xfer in the Wireless TCP optimisations
document and then the Mobile Multimedia Drafting Committee (MMDC) charter
document
from
If it is not this then still nothing is wrong, it is the
nature of
operating without using the SIM memory. If a modem is
engaged with
transmitting and a receiving SMS is incoming, then I
believe it lands up
in SIM memory.
Hm. Why do you believe this is the case ? Is it
Hi,
Does anyone have a URL for the Omnitel MAM SMSC interconnect protocol ?
Or know anything about it ?
Thanx,
I'm trying to use libXML for XML parsing, but some of the functions
there allocate memory for me (char* strings) and require me to free
them. but calling gw_free() on memory not allocated with gw_malloc()
will cause a panic, and I cannot call free() directly.
Is there a simple way to solve
coarse as ideally you'd
figure out which part- contains characters to be expanded and only reduce
its
max_part_len rather than reduce all part-'s.
Comments ?
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle Yard
Have you ever talked about montioring Kannel services with Traps ??
Its been mentioned a few times. You can get reasonable integration by
using a little abstraction - ie. Use generic SNMP agents that monitor
logfiles for keywords and other such tricks.
Maybe you can give us an operator
Lately we've been doing some very high capacity testing on
Kannel , and
found out some interesting stuff. mainly as queues (managed by lists)
are filling up to over a few hundreds of messages, the boxes start
thrashing.
I think this is directly related to the List implementation -
it's
I've added preliminary support to XML Post to smsbox.
I'm building the xml with octstr_append (this issue I guess it's
almost ok) but I'm processing the incoming xml with
octstr_search and octstr_copy :P
There's still some variables missing and alot of debug to be done.
I don't think it
;
}
as you can see, the Octstr *type is referenced after it is destroyed. Also,
correct the spelling
and indicate in the error text what MIME type is acceptable.
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle
yep the T68 does request SAR and drops the connection when kannel
answers to fall-back to non-SAR mode ;(
Maybe this is a firmware thing, but both the GTR and TTR flags are set to
1 in the WTP RcvInvoke I get from the T68 !
My firmware version is R2A013 - whats yours ?
for a specific keyword ?
Also, on a related note, the pass_data_headers() function looks for headers of
type Content-Transfer-Encoding, Content-ID and Content-Description. These are not
listed among the generic headers in 6.2.1 (WAP-145-PushMessage), so why are they
expected ?
Paul Keogh
ANAM
that the HTTPClient ptr is non-NULL before
invoking http_send_reply ().
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle Yard, Saint Patrick's Road, Dalkey, County Dublin, Ireland
that the AC_CHECK_LIB macro supports conditional actions.
Is this correct ?
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle Yard, Saint Patrick's Road, Dalkey, County Dublin, Ireland
is wrong.
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle Yard, Saint Patrick's Road, Dalkey, County Dublin, Ireland
Kannel currently has configuration variable maximum-queue-length. See
user-
guide for details.
Yes, but when this is triggered bb_smscconn_receive () logs the event and
returns -1. All the SMSC drivers except HTTP ignore the return code from
bb_smscconn_receive (). Therefore, the message
Does anyone know of any activity around the wireless TCP profile from
WAP 2.0 ? Any comments on how (if ?) this is going to go into standard OS
distributions ?
I see that CMG have announced Wireless TCP for their WAP gateway so I was
curious if wireless TCP is now available or if its still a
Maybe what is neccessary is a thread to watch over the
connection pool and recycle a connection once the server
drops it. else its a problem with the client implementation,
if it opens many connections but never use them - the
underlying connection library should not get involved in such
I'm not saying that the implementation is broken; just that
its possible to
cause descriptor starvation in a Kannel HTTP server
relatively easily using
a Kannel HTTP client.
Paul, can you provide us a szenario how we can reproduce that
behaviour? Maybe a shell script or something
And just who would execute the stylesheet? Kannel does not
include a library that provides a XSLT transformer...
I think it's a stupid idea that is hard to implement and beyond
the scope of Kannel. You can always transform the XML-data on the
client side.
Well, its not that hard to
+-+
XML Input Stream -| XSLT Engine | - XML Ouptut Stream
+-+
^
|
XSLT Stylesheet
This way the producer
Im doing some MMS experiments and using kannel as the gateway
to POST messages through. If i POST a multimedia message from
the phone that is over 900bytes kannel truncates it to
900bytes. seems to work fine with anything under 900bytes.
This is a (lack of) SAR support issue. The 900
Hi,
Has anyone tried to run the WAP g/way behind LVS ?
Thanx,
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle Yard, Saint Patrick's Road, Dalkey, County Dublin, Ireland
- Does anybody know if an Open Source client WAP stack already exist?
Not to my knowledge.
- If not, it it possible to build one from Kannel's code?
Of course. Looking at the PPG might be a good place and it's a WAP client
in some respects.
This may be difficult since the protocols
Does anyone know if you can OTA MMSC settings to a nokia 7650
and if not
why??
Check the current (0.7?) Nokia OTA spec. on http://forum.nokia.com
The function
static void client_shutdown(void)
should include the line;
client_threads_are_running = 0;
otherwise repeated use of http_start()/http_shutdown()
causes a crash.
Use case is reuse of Kannel HTTP client code in a COM object.
2002-08-28 14:24:02 [1] DEBUG: WSP 74: New state CONNECTED
2002-08-28 14:25:04 [10] ERROR: connect failed
2002-08-28 14:25:04 [10] ERROR: System error 110: Connection timed out
2002-08-28 14:25:04 [10] ERROR: error connecting to server
`www.d2chat.de' at port `80'
2002-08-28 14:25:04 [10]
revision 1.4
retrieving revision 1.5
diff -r1.4 -r1.5
2848a2849
octstr_append_char(packed, 0X01);
What exactly does the 0x01 value represent here ?
Thanks,
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047
we discovered this while figuring out why MMS notifications haven't
been working when using PPG to send the content.
This push WSP headers must have the Content-type header as the first
byte in the headers group.
The 0x01 means the length of the Content-type header, so 1 bytes
follows. And
We are using Kannel 1.2.0 with SAR-cumulative patch applied.
There seems to be lots of problems with newer Nokia phones.
File is to large errors are occuring VERY often.
Why does this happen??
Other phones are working fine, and as well, Nokia phones are
working very
fine together with
and
gwthread_wakeup() panic()'s on a -1 argument.
Suggested patch;
if (smpp-transmitter = 0) gwthread_wakeup(smpp-transmitter);
if (smpp-receiver= 0) gwthread_wakeup(smpp-receiver);
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired
Yes, there is indeed is a bug in wapbox (ppg) sms splitting.
You should
remove gw/wapbox.c, line
max_msgs = (msg_len/MAX_SMS_OCTETS) + 1;
Are you sure about this ? - it's working OK for me...
(wap.push.ppg, 0, PPG: send_push_response: telling pi);
---
debug(wap.push.ppg, 0, PPG: send_push_response: telling pi (%s),
octstr_get_cstr (reply_body));
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353
propose to add a separate return code for conn failures (-2)
so that the calling function can attempt to reopen the connections
subject to reopen delays and number of retry attempts etc.
So view this as a bug report and a proposed solution. I'll post
a patch once I've validated it.
Paul Keogh
ANAM
waiting for that. Is it coming in the near future?! Should be done
very simple I guess.
Here is a patch for the infinite loop problem I reported in the SMPP
driver for review. I also added in a reconnect-attempts configuration
variable for control. This is mandated by a number of operators
()
through calls to;
ERR_free_strings ();
EVP_cleanup();
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle Yard, Saint Patrick's Road, Dalkey, County Dublin, Ireland
Has anyone ever seen this before ?
2002-12-03 22:22:30 [12] DEBUG: AT2[087xxx]: -- AT+CPMS?^M
2002-12-03 22:22:31 [12] DEBUG: AT2[087xxx]: -- +CPMS:
SM,2,25,SM,2,25,SM,2,25
2002-12-03 22:22:33 [12] DEBUG: AT2[087xxx]: -- OK
2002-12-03 22:22:33 [12] DEBUG: AT2[087xxx]: 2 messages
Haven't encountered this, but it should be no problem
Thats good, but it should not log an ERROR for something
that's no problem. I suggest that if the AT+CMGR returns
0 length, then no PDU decode is attempted.
attachment: winmail.dat
;
_init ()
_shutdown ()
_init ()
_shutdown ()
within a single process, every dies on the second _init ().
Does'nt affect the Kannel processes directly, but is important for
anyone using the libraries elsewhere.
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353
(dlr-source);
whereas the MySQL one does not;
msg-sms.sender = octstr_duplicate(source_addr);
msg-sms.receiver = octstr_create(dst);
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless
));
octstr_destroy (account);
}
}
1033c1043
dlr_mask, pid, alt_dcs, rpi, smsc);
---
dlr_mask, pid, alt_dcs, rpi, smsc, account);
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]]On Behalf Of Stipe Tolj
Sent: Thursday, January 02, 2003 3:01 PM
To: [EMAIL PROTECTED]
Subject: Re: [RFC] new development release
I see currently still an open issue that may be a release
show-stopper.
It's
After browsing you can select from menu something like 'Quit
browsing' and
user agent correctly sends disconnect event. But I suspect if
you simply
press HangUp button, UserAgent simply go out without disconnect event.
(I checked it at least with some phones).
And users mostly end its
Hi, list.
There is bug in wml_compiler.c that forces wapbox panic
in some circumstances.
Problem is in string table.
string_table_add() returns unsigned long offset for new entry and
after that wml_compiler uses output_st_char()-octstr_append_char() to
add this offset to binary wml. It adds
);
if (conn) counter_increase(conn-sent);
Thus, conn-sent is incremented twice. The counter_increase() in the
AT2 driver should be removed.
Paul Keogh
ANAM Wireless Internet Solutions
http://www.anam.com
+353 1 284 7555 (wired) +353 86 234 6047 (wireless)
Castle Yard
This patch solves smsbox crash problem in
urltrans_get_pattern() (urltrans.c).
Not every SMSC (i.e. CIMD2 and OISD) return additional text message
in DLR. So, when this text is NULL smsbox will crash.
Best regards,
Dariusz Markowicz
+1 for this - octstr_split_words() cannot be passed
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]]On Behalf Of Maria Turk
Sent: Tuesday, February 04, 2003 11:41 AM
To: '[EMAIL PROTECTED]'
Subject: MMS binary
Does anyone have an mms binary file that they know works with
GX10 or T68i.
Or if anyone know
This isn't Kannel's decision to make; this is data set by the
end user's
telephone, which may have requirements neither you nor I are
(or could
be) aware of. The analogy to a case-insensitive filesystem
namespace is
quite fitting if you think about it carefully.
I did look at
I'm looking for some information about configuring an OpenWave browser using
e.g. push.
Nokia/Ericsson has the OTA specification describing how to setup their
browser using wsp unconfirmed push. Does something similar exist for the
OpenWave browser? (without having to use an OpenWave GW)
71 matches
Mail list logo
