Re: multiple bearer box ?
Citando Stipe Tolj [EMAIL PROTECTED]: Asif Ali wrote: hello all, i am new to this mailing list, i want to know if anybody's working on multi-bearer box architecture or there are any plans as it was identified as one of the key performance bottleneck issues, also the persistance of message streams as they pass through bearer box. hmm, I had that discussion with Alex from Centrium. They run an one bearerbox per smsc link architecture. It is definetly more performative if you get rid of all the required routing inside bearerbox if you use a bunch of smsc links inside one bearerbox, but it's more of theoretical interest. We run over 80 links in one bearerbox and fully stable. BTW, the performance bottle-neck will be of significanse when you get beyond 400 Msg/sec. (!) in MT/MO traffic, which I honestly don't expect you to have ;) We have only one bearerbox and one smsbox for 16 connections and alot of messages per day, and sometimes I'm scared when I enable a 50msg/sec on one of those connections for doing mass spam but then, while I'm sending them, I'm watching and testing the other connections and kannel works like a charm. Please note that usually I use post-xml to enable me to send like 10k destinations at one time, and thus have smsbox processing a xml post with 350KBytes, have the xml processing overhead, and have one smsbox thread creating that many messages in queue. What I mean it's might be a bigger peak processing than having 10k GET's and it works perfectly Stipe [EMAIL PROTECTED] --- Wapme Systems AG Vogelsanger Weg 80 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 13:27:05 up 88 days, 14:42, 2 users, load average: 0.14, 0.15, 0.09 'Linux is obsolete -- Andrew Tanenbaum'
Re: multiple bearer box ?
Citando Stipe Tolj [EMAIL PROTECTED]: What was the highest peak (msg/sec.) you had in the bearerbox (over all connections)? I think my peak was in a campaign that was delayed and then I needed every resource available to do it and I had something like 50+15+5+3+2+several 1 msg/sec connections, which gives like 80 msg/sec -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 20:36:09 up 88 days, 21:51, 4 users, load average: 0.85, 0.54, 0.36 'AUTHOR FvwmAuto just appeared one day, nobody knows how. -- FvwmAuto(1x)'
Re: Cookie handling
Citando Stipe Tolj [EMAIL PROTECTED]: Bruno Rodrigues schrieb: BTW: has anyone done anything for gw authentication ? I'm thinking about doing code to enable user/pass authentication in wap gw and have a exec/get-url/post-url passing user/pass and expecting a msisdn. Then have something like map-url to define which url's will wap gw send the msisdn, and which header name will it have. the WAP gateway supports HTTP Basic Authentication as specified in RFC2616. Same with get-url. It uses the get-url = user:pass@http://www.foobar.com; syntax. I mean authentication at mobile side - in your mobile configuration, you have credentials for csd(ras) or gprs(apn) authentication, which you'll use the one you get from your mobile provider, and a gateway authentication which will be sent to kannel. I intent to use the gateway credentials to translate user+pass+sourceip to msisdn and then have kannel injecting an msisdn header at the http side. When kannel receives a new connection (from a new user+pass+ip), it will request an external script/url with those values. this script/url will give back a msisdn or a reject signal. Then, when kannel fetch content from external servers, it will include the msisdn header. For that I'll also need to code an url list of can-send-msisdn and code to implement several known msisdn headers: for example, cmg is a regular header and ericsson uses a packed msisdn value. Although we can have an example script in kannel cvs, you'll need to do your own script to connect to your local architecture - it could be a script querying a radius server or a script authenticating against mysql/ldap/SSO system. -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 20:38:06 up 88 days, 21:53, 4 users, load average: 0.57, 0.54, 0.37 'finlandia:~ apropos win win: nothing appropriate.'
Re: T68i incorrect support of CNMI
Citando Alex Judd [EMAIL PROTECTED]: Yes - it does look like it's well and truly broken. However... it does work fine if it's configured with sim-buffering so perhaps we change the default settings for the T68i to be that. At least that way it works? Yes please, just edit modems.conf and send us the patch. If you want, you can even add comment lines # 3,1,2,3 this should work because manual says so but this stupid mobile simply rejects us. ;) fortunatly, it's a great small mobile to carry inside your pocket and feed your pda with gprs and bluetooth - If i have 3 timeslots available, t68 will deliver me the full 43.2kbps :)) but for calling or messaging, forget Alex - Original Message - From: Bruno Rodrigues [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Friday, February 14, 2003 8:42 PM Subject: Re: T68i incorrect support of CNMI Citando Alex Judd [EMAIL PROTECTED]: RE: [Fwd: Removing WTP-SAR limit 32768 B]Anyone else noticed that the T68i appears to break the rules of the CNMI command in smsc_at2.c? AT+CNMI=? +CNMI: (3),(0,1,3),(0,2),(0),(0) Haven't I commented modems.conf with something like # Ericsson T68 - forget it ? -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 09:55:17 up 87 days, 11:10, 7 users, load average: 0.06, 0.03, 0.00 'World domination. Fast (By Linus Torvalds)'
Re: Billing Kannel SMS side
Citando Arne K. Haaje [EMAIL PROTECTED]: I am thinking about adding support to the CIMD2 module for the Tarrif field. This is in Nokias official specifications. It is a field like any others where you can set the tarrif class. As it is in the official specs. I think it would be useful. Are there any objections to me adding support for it? If not, would the right way be to let the user specify it in the URL for sendsms - tarrif=12 ? UCP 4.0 5.1.2.5 XSer Type of service 0C: Billing Identifier This type of service enables Large Accounts to send additional billing information to the SMSC. Billing Identifier data element is an alphanumeric field with a variable length of at least 0 and at most 20 characters. These characters need to be part of the Visible String character set as defined in ITU-T. Each character takes two hexadecimal positions. I need a 0-20 chars parameter for UCP :) tarrif or billing is ok for me. -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 10:09:34 up 87 days, 11:25, 7 users, load average: 0.00, 0.04, 0.00 'Being overloaded is the sign of a true Debian maintainer. -- JHM on #Debian'
Re: Rotatelog for traditional Unix systems
Citando Alex Judd [EMAIL PROTECTED]:
Everyone
Additional comments on rotating logs for the userguide for systems that
don't have RedHat's logrotate installed.
+ ACTIONS:
+ kill - HUP | ps -ef | grep -i production | awk '{print $2}' :
/var/log/kannel/access.log, /var/log/kannel/enpocket.log,
/var/log/kannel/smsbox.log
+ NOTIFY:
+ [EMAIL PROTECTED]
kill - HUP | something ? what's this ? shouldn't it be something like
kill -HUP `something | otherthing` ?
isn't there a pidof command ?
and won't rotatelog support /var/log/kannel/*.log ?
--
Davi / Bruno.RodriguesatLitux.Org
Litux.org: 10:29:55 up 87 days, 11:45, 7 users, load average: 0.12, 0.07, 0.01
'Turn right here. No! NO! The OTHER right!'
RE: Kannel wap limits ?
Citando Igor Ivoilov [EMAIL PROTECTED]: I have to fix wpt layer to do not panic if psn is going to be 255 but this is a limit of SAR and can be overcome with ESAR. Does 3650 support ESAR? I don't have such a phone to check it by myself So it seems altough kannel's only relevant information is the 350K sdu. -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 10:54:26 up 87 days, 12:10, 8 users, load average: 0.03, 0.05, 0.01 'Linus? Whose that? -- clueless newbie on #Linux'
Re: DCS and PID handling in Kannel
Citando Stipe Tolj [EMAIL PROTECTED]: If it's DCS is really broken up in MO SMPP, then please re-send the patch. Use [PATCH] in the mail subject line. And try to explain in the mail what the problem was and what you patch does to fix it. Thanks in advance. I'm definetly willing to review it and vote. At least we can offer you that alternative option. If you want I can re-ask for a SMPP connection and do some debugging in here to fix what's wrong with with fields = DCS. -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 12:42:06 up 87 days, 13:57, 2 users, load average: 0.24, 0.10, 0.03 'Sex dumps core (Sex is a Simple editor for X11) -- Seen on debian bugtracking'
Re: DCS and PID handling in Kannel
Citando Dave White [EMAIL PROTECTED]:
Say we have a binary (8-bit) message without compression, but with
message class 1. These are pretty common; Nokia phones use these for
most Smart Messaging MO's.
There are two valid representations of this as a DCS octet:
(WARNING -- try a fixed-width font if this looks like junk)
0001 0101 = 0x15
^^ ^ ^
|| | |_ Message Class 1
|| |___ 8 bit data
||__ Message class present
|___ No compression
0101 = 0xF5
^ ^ ^
| | |_ Message class 1
| |___ 8 bit data
|__ Message class present
Another fun example is 7-bit flash (Class 0) SMs
0001 = 0x10
= 0x240
Given the information Kannel passes over HTTP POST, it is impossible to
decide which of the two spellings of an equivalent DCS was originally
sent in the MO.
Kannel should decide for the 0x0x spelling unless you pass alt-dcs field,
which will select the 0xFx spelling.
see X-Kannel-Alt-DCS
--
Davi / Bruno.RodriguesatLitux.Org
Litux.org: 12:55:12 up 87 days, 14:10, 4 users, load average: 0.15, 0.13, 0.06
'lp1 on fire
-- One of the more obfuscated kernel messages'
Re: DCS and PID handling in Kannel
Citando Dave White [EMAIL PROTECTED]:
Say we have a binary (8-bit) message without compression, but with
message class 1. These are pretty common; Nokia phones use these for
most Smart Messaging MO's.
There are two valid representations of this as a DCS octet:
(WARNING -- try a fixed-width font if this looks like junk)
0001 0101 = 0x15
^^ ^ ^
|| | |_ Message Class 1
|| |___ 8 bit data
||__ Message class present
|___ No compression
0101 = 0xF5
^ ^ ^
| | |_ Message class 1
| |___ 8 bit data
|__ Message class present
Another fun example is 7-bit flash (Class 0) SMs
0001 = 0x10
= 0x240
Given the information Kannel passes over HTTP POST, it is impossible to
decide which of the two spellings of an equivalent DCS was originally
sent in the MO.
Kannel should decide for the 0x0x spelling unless you pass alt-dcs field,
which will select the 0xFx spelling.
see X-Kannel-Alt-DCS
--
Davi / Bruno.RodriguesatLitux.Org
Litux.org: 12:55:12 up 87 days, 14:10, 4 users, load average: 0.15, 0.13, 0.06
'lp1 on fire
-- One of the more obfuscated kernel messages'
Re: [PATCH] Fix for gw/smsc/smsc_smpp.c - bogus coding and no udh on deliver_sm
Citando Dave White [EMAIL PROTECTED]:
This patch fixes the following issues:
*) Kannel coding param is overwritten with wrong values by pdu_to_msg
(e.g. a Nokia business card, which is 8-bit data encoded as iso-8859-1,
is processed by Kannel as 7-bit text, and will be forwarded as coding 1.)
I see your point now. Setting msg-sms.coding by hand is broken. If you can
gather a dcs field from smsc pdu, you should call:
if (! dcs_to_fields(msg, dcs)) {
error(0, EMI2[%s]: invalid dcs received,
octstr_get_cstr(privdata-name));
/* XXX Should we discard message ? */
dcs_to_fields(msg, 0);
}
which will fill sms.coding, sms.compress, sms.mclass and sms.mwi correctly.
I'm now commiting a quick patch to set alt-dcs too on MO messages.
Similary, you should get msg and passing it to fields_to_dcs(msg), you'll get
the right dcs based on coding, compress, mclass, mwi and alt-dcs
*) UDH is ignored in SMS-MO
If SMPP developers agree with this code, I vote +1 for it and give a thank you
very much for your help
--
Davi / Bruno.RodriguesatLitux.Org
Litux.org: 12:59:19 up 87 days, 14:14, 4 users, load average: 0.37, 0.17, 0.08
'Linux is obsolete
(Andrew Tanenbaum)'
Re: [PATCH] Fix for gw/smsc/smsc_smpp.c - bogus coding and no udh on deliver_sm
Citando Dave White [EMAIL PROTECTED]:
This patch fixes the following issues:
*) Kannel coding param is overwritten with wrong values by pdu_to_msg
(e.g. a Nokia business card, which is 8-bit data encoded as iso-8859-1,
is processed by Kannel as 7-bit text, and will be forwarded as coding 1.)
I see your point now. Setting msg-sms.coding by hand is broken. If you can
gather a dcs field from smsc pdu, you should call:
if (! dcs_to_fields(msg, dcs)) {
error(0, EMI2[%s]: invalid dcs received,
octstr_get_cstr(privdata-name));
/* XXX Should we discard message ? */
dcs_to_fields(msg, 0);
}
which will fill sms.coding, sms.compress, sms.mclass and sms.mwi correctly.
I'm now commiting a quick patch to set alt-dcs too on MO messages.
Similary, you should get msg and passing it to fields_to_dcs(msg), you'll get
the right dcs based on coding, compress, mclass, mwi and alt-dcs
*) UDH is ignored in SMS-MO
If SMPP developers agree with this code, I vote +1 for it and give a thank you
very much for your help
--
Davi / Bruno.RodriguesatLitux.Org
Litux.org: 12:59:19 up 87 days, 14:14, 4 users, load average: 0.37, 0.17, 0.08
'Linux is obsolete
(Andrew Tanenbaum)'
Re: DCS and PID handling in Kannel
Citando Dave White [EMAIL PROTECTED]: Bruno Rodrigues wrote: Kannel should decide for the 0x0x spelling unless you pass alt-dcs field, which will select the 0xFx spelling. 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 alt-dcs, and I consider it a kludge. Sorry. It's not a Kannel decision, it's a kannel option did you understand the you can generate every valid dcs ? did you understand that third parties, which might not have a clue about dcs and like to just say mclass=1 for flash or coding=3text= for unicode accented chars? sorry. it's not my decision anymore. I vote +0 for a dcs field but no changes to other fields. -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 13:19:04 up 87 days, 14:34, 4 users, load average: 0.06, 0.04, 0.04 'Turn right here. No! NO! The OTHER right!'
Re: Cookie handling
Citando Wilms, Stefan, VIS-THND [EMAIL PROTECTED]: Dear all, trying to get access to our intranet via WAP, using kannel 1.1.6 (with --enable-cookies compiled) There's something wrong with cookie support but as I'm gonna need them to work soon for a project of mine, I'll look at the code and will make sure they work. -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 19:07:34 up 87 days, 20:23, 5 users, load average: 0.07, 0.09, 0.16 'May you do Good Magic with Perl. -- Larry Wall's blessing'
Kannel wap limits ?
Hi.
I'm trying to force kannel to let a mobile download realplayer.sis file
(584.720bytes) through it.
I know that I'm crazy to try it, but I just need to know the limits of kannel
and see if we can do better than CMG ;)
First problem is that Nokia 3650 sends a SDU of 357.000 bytes and kannel simply
refuses to send the file back.
It's a good start relativly to CMG, because I will get a reply quickly (altough
it could be a nice wap page saying something meanfull instead of a rude gateway
error). CMG tries to do something and only breaks after something like 10 seconds.
Then I've removed SDU checks from kannel :), adding a 0
gw/wap-appl.c line 654:
/*
* If the response is too large to be sent to the client,
* suppress it and inform the client.
*/
if (0 octstr_len(content.body) sdu_size sdu_size 0) {
Now the mobile goes having fun downloading the file:
Kannel logs starts with a:
2003-02-18 02:52:12 [5] DEBUG: WTP: begin_sar_result(): data len = 584867
then a,
2003-02-18 02:52:14 [5] DEBUG: WTP: continue_sar_result(): lsegm=2, nsegm=1015,
csegm=-1
until it gets to:
2003-02-18 02:54:24 [5] DEBUG: WTP: resp_machine 2, state RESULT_RESP_WAIT,
event RcvAck.
2003-02-18 02:54:24 [5] DEBUG: WTP: continue_sar_result(): lsegm=254,
nsegm=1015, csegm=251
2003-02-18 02:54:24 [5] DEBUG: WTP: dispath_to_wdp(): psn = 255
2003-02-18 02:54:24 [5] PANIC: gwlib/octstr.c:1677: octstr_set_bits: Assertion
`value = mask' failed.
This means what ? are we assuming that psn is a unsigned byte ?
To reach the Nokia3650 SDU, we would need to send 620 psn's, so I think
something is broken in kannel. :(
Any idea ?
--
Davi / Bruno.RodriguesatLitux.Org
Litux.org: 02:58:48 up 87 days, 4:14, 10 users, load average: 0.18, 0.10, 0.05
'OK, enough hype.
-- Larry Wall in the perl man page'
Re: DCS and PID handling in Kannel
Citando Dave White [EMAIL PROTECTED]: The GSM data coding scheme (DCS) is broken up by Kannel into various seperate fields, as I've seen. The protocol id, however, is passed through as an opaque octet. Why is that? PID could concievably be treated the same way as DCS, and broken up into human-friendly fields like replace_msg_class, message_type, etc. On the other hand, I'd actually prefer the following: DCS and PID should also be accepted (and sent to applications) as-is; any additional fields are sugar for the application (so the developers of CGI servers for Kannel don't need to deal with the full hairiness of the GSM params). My problem with Kannel's DCS handling is that there are in many cases two valid representations for certain combinations of message class and user data coding scheme. Since Kannel shreds DCS into a bunch of fields, if I want to reconstruct the original DCS of the MO SM, I have a harmless but icky ambiguity; in many cases, I'll provide a new value that SHOULD be semantically equivalent, but MIGHT not be for some buggy devices/firmware/etc. Wouldn't it be easier just to add an INTEGER field to the sms message type, and pass the DCS through as-is? PID value is passed thru from smsbox http interface to smsc connection and kannel doesn't mind with it. DCS is used internally in kannel and thus should be controled by it. That's why we have seperated fields to build DCS at the end. For example, if you set UDH and don't set coding, kannel will automatically (legacy behaviour) select 8 bit text and adapt your text data to it. Using coding, mclass, mwi and alt-dcs, every dcs combination can be done, if smsc modules supports it of course. EMI2 and AT2 surelly supports it. I can make a table in userguide with it's combinations to describe it. I vote -1 for having a DCS field outside smsbox and I don't see a requirement to split pid, but that's only MHO :) Thanks, David WHITE CONNECT AUSTRIA -- Davi / Bruno.RodriguesatLitux.Org Litux.org: 06:05:56 up 87 days, 7:21, 16 users, load average: 4.23, 3.66, 2.65 'Linux is obsolete (Andrew Tanenbaum)'
Re: [Fwd: Removing WTP-SAR limit 32768 B]
Citando Stipe Tolj [EMAIL PROTECTED]: Igor, Igor Ivoilov schrieb: the same problem for me with your implementation ;((( I've tried FirstHop gw, it has the same problem with GX10 as Kannel can you describe what happens while the problem appears? AFAIK, Bruno was also experimenting with the Sharp thingy. I don't have now a GX10 at my hands to test it but I confirm that that's an error from mobile, altough I don't recall it being at 32k. As a matter of fact, watching some internal logs ;) I see some jar files with 50KB. I guess our mobiles do handle more than 32k, but less than 90K, because I recall that jabber_midp.jar with 99.502 gaves me that kind of error but some java games I have in here have 50KB each. I'm now trying to download jabber_p800.jar (99.572bytes) with a Ericsson P800 using kannel 1.3.0 and it worked perfectly (and full speed) -- br/
Re: receiving a Vcard and save toa DB
Citando Stipe Tolj [EMAIL PROTECTED]: joseph mercado wrote: Has anyone already implemented kannel to recieve a Vcard and point to a url where it get parse and be save to DB I have played a bit with the Nokia smart messaging stuff. I didn't recall exactly if the vcard is delivered with alternative DCS fields?! AFAIK, binary DCS field are not supported for most SMSC modules. *ehmmm* *ducking head* Stop complaining and implement it ;) Hope others that run such services don't flame me now ;) VCard's are in simple text. Nokia phones requires a special UDH. Ericsson's don't. If you want to send, set udh=%06%05%04%23%F4%00%00 for nokia and empty udh for ericsson, and set text=see example downthere If you want to receive, you'll need a catchall service that forwards %a to your server, and you can ignore udh field - that way you'll receive from nokia and ericsson. Your server needs to parse vcard and reply back acordly. Please have a look at contrib/web/*.php Example vcard: (Note: with old nokias, like 6210, you can only set maximum two fields from email, label and note. If you set the three, it will ignore your message. I haven't tried with new mobiles) Note2: newlines are %0D%0A, and you need a newline after last line (END:VCARD%0D%0A) BEGIN: VCARD VERSION:2.1 N: name TEL;PREF: telephone TEL;CELL: mobile TEL;HOME: home phone TEL;WORK: work phone TEL;FAX: fax EMAIL: email LABEL: address NOTE: note END:VCARD But the idea is cool. Stipe [EMAIL PROTECTED] --- Wapme Systems AG Vogelsanger Weg 80 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are -- br/
Re: Urgent Query ??
Citando Stipe Tolj [EMAIL PROTECTED]: No, Kannel is not fully WAP 1.2.1 compliant. We cover the most important parts: WTP-SAR, WAP Push, no UAProf. I might have some news about UAProf one of these days ;) Hint: Deli, Cocoon, Vodafone live -- br/
Re: T68i incorrect support of CNMI
Citando Alex Judd [EMAIL PROTECTED]: RE: [Fwd: Removing WTP-SAR limit 32768 B]Anyone else noticed that the T68i appears to break the rules of the CNMI command in smsc_at2.c? AT+CNMI=? +CNMI: (3),(0,1,3),(0,2),(0),(0) Haven't I commented modems.conf with something like # Ericsson T68 - forget it ? I've tried every combination and the I've given up. Ericsson reply was won't fix it :((( -- br/
Re: Fw: kannel dependancies
Citando Derry Crittenden [EMAIL PROTECTED]: Sorry to bother you, problem solved by installing 1.2.0 Next time you release something, please consider what packages it really requires. Hi there, Since kannel version 1.2.1-2 conveniently requires versions of glibc6, libpam0g and libxml2 greater than those on the australian debian mirror, aarnet.edu.au and the debian 3 r.1 dist cd, can you please suggest somewhere i might find the packages required to install it. I've updated kannel homepage and I've included kannel-1.2.0 packages for woody and sarge distribution, which should resolve your problem. After Stipe releases devel 1.3.1, I'll generate and upload new debian packages for that release. I don't have time now for generating 1.3.0 packages when we are so close to 1.3.1 release. -- br/
Re: Text messages with UDH, bad DCS
Citando Enzo dell'Aquila [EMAIL PROTECTED]: We do use UDH 8-bit messages over EMI2 links that do work as expected. I don't know how emi2 link works but i need send a sms with DCS = 00 (or 11 or 12) [GSM 7 bit default alphabet], body 7bit encoded and UDH 8 bit encoded. I know this isn't possible with other drivers, I can't obtain this DCS with kannel. Is this a kannel problem or a emi2 problem? A lot of sms need this configuration! have you seen my mail with kannel GET parameters to produce those DCS's ? Enzo dell'Aquila -- br/
Re: Kannel cvs problem
Citando Alan McNatty [EMAIL PROTECTED]: Thanks Bruno, I did a cvs diff across 1.2.1 vs 1.3.0 to find ... ./.cvsignore 5,6d4 config.guess config.sub 9,11d6 diff diff2 build-stamp looks like someone has removed the config.sub and config.guess files from cvs repository - these are required as part of the build process. [ ie: if I copy them from 1.2.1 source bundle things compile nicely ]. I did it. Both files are generated when you run autoconf in you kannel directory. I've deleted them after seeing that debian package building system have replaced them by new ones and, after built the package, lintian complain about these two automatic generated files inside it. They are in cvs again. Cheers, Alan On Mon, 2003-02-03 at 16:26, Bruno Rodrigues wrote: Quoting Alan McNatty [EMAIL PROTECTED]: I suspect there is a problem with current cvs. ie: [ clean checkout ] ... gateway$ ./configure loading cache ./config.cache Configuring for Kannel gateway version 1.3.1 ... Running system checks ... configure: error: can not run ./config.sub I'll fix that in some minutes. -- br/ - This mail sent through IMP: http://horde.org/imp/
Re: Where's my FOO ? [Was: Daily patch: gateway]
Citando Konstantin Vayner [EMAIL PROTECTED]: Hi, Actually... What is really the reason for not having really descriptive status text? There is already a list of definitions for status-codes... So i believe it would not be a problem to make same for status-text... or is it? Having OK as status-text is really confusing... but so also having Foo/Bar in some cases... So... [ if i may really vote on anything ;) ] i would vote +2 on making status-text be real rfc compliant (and thus descriptive) or +1 for having at least Foo/Bar instead of OK Are you offering to build that table in .c for kannel ? ; -- br/ - This mail sent through IMP: http://horde.org/imp/
RE: EMI Driver Bug
Citando Michael Mulcahy [EMAIL PROTECTED]: Hi All, Comments Below: You're mixing up the SMSC response with the delivery report. When you send a message, you send a type 51 message. you will get a 50ACK back saying the SMSC has accepted the SMS. Is this a typo on your part? According to the EMI spec 4.0 the response to the submit short message operation 51 should contain an operation type of 51 not 50. Indeed 51/R This is NOT the delivery report. The SMSC response is acknowledging that the SMSC has accepted the message but it doesnt say that the message has been delivered to the handset. I think there is a misunderstanding of terminology here. I use the term submission notification to mean that the message was submitted to the SMSC not delivered to the handset. I guess you are right, although old versions didn't have this problem - I use dlr=255 to activate every kind of dlr in my some-month-old kannel in production. This code handles only receiving and 'R' packets. It's an if( OR=R OT=51) then, so there's no 50_AMSG field available. Then there's a different code to handle OR=O, OT=53 at line 721, and there is an if((emimsg-fields[E50_AMSG]) == NULL) msg-sms.msgdata = octstr_create(Delivery Report without text); else msg-sms.msgdata = octstr_duplicate(emimsg-fields[E50_AMSG]); This looks like a copy-paste to me. Your patch is cleaner that doing something like this lines because, indeed, there's no 50_AMSG in OR=R I'm +1 for it Look forward to hearing from you, Warm Regards, Michael. ANAM Wireless Internet Solutions http://www.anam.com mailto:[EMAIL PROTECTED] +353 1 284 7555 Castle Yard, Saint Patrick's Road, Dalkey, County Dublin, Ireland -Original Message- From: Andreas Fink [mailto:[EMAIL PROTECTED]] Sent: 03 February 2003 13:19 To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: EMI Driver Bug On Montag, Februar 3, 2003, at 01:23 Uhr, Michael Mulcahy wrote: Hi All, Summary: The assigning of a nonexistent field in the EMI driver when creating a submission report results in an out of bounds read. Scenario: The EMI driver checks for a DLR when a response is received for a submitted message. If there is a DLR requested for that message then the driver does the following: /* * Recode the msg structure with the given msgdata. * Note: the DLR URL is delivered in msg-sms.dlr_url already. */ dlrmsg-sms.msgdata = octstr_duplicate(emimsg-fields[E50_AMSG]); octstr_hex_to_binary(dlrmsg-sms.msgdata); dlrmsg-sms.sms_type = report; Why does the driver assign the value of the E50_AMSG field to the msgdata of the dlr message? The AMSG field does contain the delivery report text of the SMSC. This is a text of style The message to 12345 with referenfce number 1847127 has been delivered on 12.1.2003 14:25 This field is not available in the EMI response. The response EMI message only has three fields so the above code accesses data beyond the array bounds as E50_AMSG has a value of 20. You're mixing up the SMSC response with the delivery report. When you send a message, you send a type 51 message. you will get a 50ACK back saying the SMSC has accepted the SMS. This is NOT the delivery report. The SMSC response is acknowledging that the SMSC has accepted the message but it doesnt say that the message has been delivered to the handset. When the message has been delivered to the handset, the SMSC generates an incoming message of type Delivery Report which has all fields, much similar to an incoming SMS. Andreas Fink Global Networks Switzerland AG -- Tel: +41-61-333 Fax: +41-61-334 Mobile: +41-79-2457333 Global Networks, Inc. Clarastrasse 3, 4058 Basel, Switzerland Web: http://www.global-networks.ch/ [EMAIL PROTECTED] -- Member of the GSM Association -- br/ - This mail sent through IMP: http://horde.org/imp/
Re: MMS files.
Citando Maria Turk [EMAIL PROTECTED]: Does anyone know if there is a difference between MMS formats between nokia 7650 sony ericsson T68i and the GX10. Yes, they are. But there's a common mms format that works in them all (including Panasoinc GAD87). My experience says that it's too much dependent of your mmsc, specially through email interface. You need to try to send and try with combinations of start tag and content-id Thanks, Maria Turk Applications Specialist Wizcom(Singlepoint4u) Tel:+44(0)1782600450 Mob: +44(0)7775837460 -- br/ - This mail sent through IMP: http://horde.org/imp/
Re: Where's my FOO ? [Was: Daily patch: gateway]
Citando Konstantin Vayner [EMAIL PROTECTED]: Bruno Rodrigues wrote: Citando Konstantin Vayner [EMAIL PROTECTED]: Hi, i would vote +2 on making status-text be real rfc compliant (and thus descriptive) or +1 for having at least Foo/Bar instead of OK Are you offering to build that table in .c for kannel ? ; Yes... that is just an array of strings, nothing more ;) [ and i would put it in .h - where the definitions for http-status are ;) ] Nice : Can I still have my octstr_send_random_foo_sometimes() ? :))) or a if(user-agent =~ /ms.xmlhttp/) do array[202] = I therefor accept your message but I don't fscking know if I'd want to sent it ? (kidding) Note: i am not offerring to make kannel parse these values from RFC although it is also a possability ;-)) No need for that. Status is the one that matters, status text doesn't matter at all. Regards, Konstantin Vayner Appcell MT -- br/
Re: Text messages with UDH, bad DCS
Citando Enzo dell'Aquila [EMAIL PROTECTED]: I tried to send text messages with UDH with emi2 driver. This feature appears as y? (driver probably has this feature, has not been tested). Weird. I've fixed it now in cvs. EMI2 supports everything. Unless.. I would like generate TP-DCS = 00h (it's good also 11h or 12h), I need 7 bit coding in DCS Is this a bug in emi2 kannel driver or doesn't emi2 support text messages with UDH ? In my kannel instance from Aug 2002, requesting ...text=testudh=%01% 00coding=1mclass=2 gives me DCS=11 text=testudh=%01%00coding=1mclass=0 results in DCS=00 If this don't work with you, please say so because there might be some new bug and someone will have to fix it :| -- br/
Re: Does Kannel log support logrotation
Citando Alan McNatty [EMAIL PROTECTED]: This might be useful for you ... (add to /etc/logrotate.d or similar). Note: you need to restart Kannel as it doesn't have full signal support. you just need to killall -HUP bearerbox smsbox wapbox. -HUP doesn't re-read configuration, but it reopens log files. And yes, logrotate is the way to go On Fri, 2003-01-31 at 18:18, joseph mercado wrote: Does Kannel log support logrotation???Please advice regarding this matter cause im running kannel-1.2.1 on linux box 7.2 that forwards sms to a EMI2 server.Currently all logs does'nt have rotate mechanismhas anyone implemented this already __ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com -- Alan McNatty [EMAIL PROTECTED] -- br/
Re: MMS binary code..
Citando Maria Turk [EMAIL PROTECTED]: Does anyone have any code that produces the .mms file. I tried to Nokia version but it's not that conformant.. doesn't work on many phones. This one should work with Nokia, Sharp and Ericsson. I've lost all my email and I've already asked for the mail with panasonic changes, but I think it's something like having 1. boundary without spaces and a start tag boundary=---X; start=s.smil 2. in smil, having an header Content-ID: s.smil 3. have in every content part an aditional Content-ID with the name inside Content-Type: ; name=foo.jpg Content-ID: foo.jpg 4. Content-Disposition shouldn't be required Date: Tue, 4 Feb 2003 22:00:17 + (WET) From: ... To: ... Subject: test Mime-Version: 1.0 Content-Type: multipart/related; boundary==_Part_191_16701175.1044396017841 --=_Part_191_16701175.1044396017841 Content-Type: application/smil; name=mail.smil Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename=mail.smil smil head layout root-layout / region id=Image height=100% width=100% left=0% top=0% / /layout /head body par dur=5000ms audio src=Telephone.amr / img src=BI_BG_2_MMS_7650.jpg region=Image / /par /body /smil --=_Part_191_16701175.1044396017841 Content-Type: audio/amr; name=Telephone.amr Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename=Telephone.amr snip base64 --=_Part_191_16701175.1044396017841 Content-Type: image/jpeg; name=BI_BG_2_MMS_7650.jpg Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename=BI_BG_2_MMS_7650.jpg snip base64 --=_Part_191_16701175.1044396017841-- -- br/
Re: PPG: wap-push-user specific smsc-id config
Citando Andreas Fink [EMAIL PROTECTED]: On Montag, Februar 3, 2003, at 12:52 Uhr, Stipe Tolj wrote: Hi list, Hi Aarno, a) if smsc=... is set on the push, use it otherwhise use the one defined in group=ppg b) if group wap-push user has a default-smsc-id, use it if smsc=.. is not set by step a) c) if group wap-push user has a smsc-id, discard any already set smsc-id and force it to the value of the config file. Don't we have a forced-smsc-id in sendsms ? please use it instead of smsc-id to maintain consistency ;) -- br/
Re: Kannel cvs problem
Quoting Alan McNatty [EMAIL PROTECTED]: I suspect there is a problem with current cvs. ie: [ clean checkout ] ... gateway$ ./configure loading cache ./config.cache Configuring for Kannel gateway version 1.3.1 ... Running system checks ... configure: error: can not run ./config.sub I'll fix that in some minutes. -- Alan McNatty [EMAIL PROTECTED] -- br/ - This mail sent through IMP: http://horde.org/imp/
Re: Kannel cvs problem
Quoting Alan McNatty [EMAIL PROTECTED]: I suspect there is a problem with current cvs. ie: [ clean checkout ] ... gateway$ ./configure loading cache ./config.cache Configuring for Kannel gateway version 1.3.1 ... Running system checks ... configure: error: can not run ./config.sub I'll fix that in some minutes. -- Alan McNatty [EMAIL PROTECTED] -- br/ - This mail sent through IMP: http://horde.org/imp/
Re: [PATCH] cimd2 validity period
On Sáb, 2003-02-01 at 12:21, Nicholas Rahn wrote: On Sat, 2003-02-01 at 04:47, Bruno David S=?ISO-8859-1?B?aW31ZXMgUm9kcmlndWVz?= wrote: Citando Nicholas Rahn [EMAIL PROTECTED]: Hi list, Attached is a patch implementing the validity period (relative) for CIMD2 connections. It has been tested with one of our swiss CIMD2 connections. The documentation for the validity parameter (sendsms interface) in the User's Guide does not mention that this functionality might not be implemented for certain SMSC types. A quick search for validity shows: Can set Validity ? ? ? ? y ? ? ? ? y ? ? We can't mention for each parameter that this might work or might not work for depending on connected smsc. Better solution is to ask smsc developers to implement it and fix documentation ;) Best solution is for me to actually read that section of the documentation before complaining. ;-) No, it's good if someone complains. That way someone would remember that he can do that patch and update documentation :) nick -- Bruno Rodrigues [EMAIL PROTECTED] signature.asc Description: This is a digitally signed message part
Re: EMI does not ACk message send at once
On Sáb, 2003-02-01 at 05:50, Cipher Strength wrote: send messages with some milliseconds delay to kannel. That's not a solution. Just ask your smsc provider which flow-control they are using - stop-and-wait or windowing - and if it's windowing, how which size it have CIPHER Citando joseph mercado [EMAIL PROTECTED]: Hi list; Im trying to push a very long message to EMI2 server. after the request has been answered by external API my smsbox.log got the message and split it into 3.Same things happened to the bearebox.log...it recieved 3 sms message and tries to push it on SMSC ...problem is i only get 2 smswhats wrong ???did i miss something is bearerbox sending three messages to smsc and only receiving two ack ? This looks like your smsc having a small windowing or even a window value of 1. You should set flow-control = true. Basically, smsc windowing is how many messages you can send before receiving an ack. If smsc have window=3 you can send MSG1 to 3 and then you'll have to wait for ACK1 before sending MSG4. If you send any more messages, they'll be discarder. flow-control = true is like window=1 - kannel only sends another message after receiving ack from last one. -- BR/ _ The new MSN 8: advanced junk mail protection and 2 months FREE* http://join.msn.com/?page=features/junkmail -- Bruno Rodrigues [EMAIL PROTECTED] signature.asc Description: This is a digitally signed message part
Re: multiple sms using EMS standard
On Seg, 2002-12-09 at 09:40, Dziugas Baltrunas wrote: Hello, i would like to ask, is it possible to send two or free sms within a sms-service group. currently, i'm interested in melody sending. the problem is that using nokia binary format it is possible to encode 130 bytes of data and usually melody fits in such size. but whenm using EMS, there is a limitation of 128 bytes in order to encode and to send melody in iMelody format. (...) but EMS allows to send multiple SMS using ceratin directives with UDH. i would like to ask, is it possible to do such things with kannel. if not, i think i would be interested to develope such thing. SMS Spliting with EMS should be done in application. It's too hard to be done in kannel, unless we implement the full EMS specification into it. Then, to send multiple sms's, you should use the UDH field to say how much parts and which message part you are sending. You should look at 00 UDH Function, and do something like: %00%03%00%02%01 %00-UDH Funcion 00 - Multiple messages %03-length %00-Message number (use, for example, a random for your message and should be the same value for each message part) %02%01 - one is the maximum parts, the other is the part number (I never remember which is each, but I guess it's MAX, Part number). So, you'll send something like (assuming it's max parts first) http://.sendsms?...udh=%XX%00%03%00%02%01% where XX is 5 + your message splitted part size regards, Dziugas Baltrunas -- Bruno Rodrigues [EMAIL PROTECTED] signature.asc Description: This is a digitally signed message part
Re: Bug in wtp_pack_sar_ack
On Sex, 2002-12-13 at 09:45, Aarno Syvänen wrote: Hi List, in wtp_pack_sar_ack we have following: unsigned char cpsn; sprintf(cpsn, %c, psn); Isn't this like putting an char (%c) inside a char (cpsn) with the psn value ? like cpsn = (char) psn; ? -- Bruno Rodrigues [EMAIL PROTECTED] signature.asc Description: This is a digitally signed message part
Re: OTA MMS Settings
On Dom, 2002-12-08 at 09:22, Andreas Fink wrote: On Sonntag, Dezember 8, 2002, at 01:30 Uhr, Bruno Rodrigues wrote: Yes, they do. We are provisioning MMSC OTA to Ericsson T68 and every Nokia that supports MMS (3510, 7650, 7210, etc). I can send some examples if required. yes. we're interested in examples. Are they the same for different models? As I said before, I'm using my own php scripts to build the ota binary (the ones I've commited to contrib). I don't remember if I've updated those with the MMSURL, but it's a small patch (just need to say to my script that mmsurl is tag 87, I guess). The MMS OTA is the same for nokia and T68. There were some issues with it. I guess I had to set MMSURL for T68 AND URL for Nokia 7650. The dump: # mms.dump 01062C1F2A6170706c69636174696f6e2f782d7761702d70726f762e62726f777365722d73657474696e67730081EA01016A0045C60801 871511034d4d5300010186071103687474703a2f2f6d6d732f736572766c6574732f6d6d730001867C1103687474703a2f2f6d6d732f73 6572766c6574732f6d6d730001C67F01871511032d000187171103687474703a2f2f6d6d732f736572766c6574732f6d6d73000101C606 0187124901871311033231332e33302e32372e3633000187146101871C11037661732e766f6461666f6e652e7074000187227001872311 0376617300018724110376617300010101 I'm returning the OTA with a perl script,in reply to a MO sms, like this: #!/usr/bin/perl ... check keyword and read the corresponding file ... open(FILE, $prefix/$msg.dump); @file= FILE; close(FILE); $file[0] = pack(H*, $file[0]); $lf = \n; $udh=060504C34F; print X-Kannel-UDH: . $udh. $lf; print X-Kannel-Coding: 2. $lf; print Content-Type: application/octet-stream.$lf; print $lf; print $file[0]; The variables I set are: # mms.data file=mms name=MMS urlname=- url=http://mms/servlets/mms mmsurl=http://mms/servlets/mms bearer=49 ppp_authtype=70 ppp_authname=user ppp_authsecret=pass proxy=1.2.3.4 port=61 gprs_accesspointname=foo.vodafone.pt The script I'm using to write the mms.data and mms.dump is attached PS: my php to send wappush si stopped working in here. I haven't changed my code, so I guess something happened to smsc configuration. Can you send me and example (kannel.conf, cmd line and xml files) to send a wappush (text + urllink) through kannel interface ? -- Bruno Rodrigues [EMAIL PROTECTED] write.php Description: application/php signature.asc Description: This is a digitally signed message part
RE: OTA MMS Settings
On Qua, 2002-12-04 at 10:46, Igor Ivoilov wrote: Sorry, I did not look into the sms content at the first glance But now I hardly can recognize is as OTA settings See OTA setting specification v.7.0 from nokia and ericsson about how to do it. This document was somewhere on the nokia site. See also http://www.3glab.org/ota.html My previous advise was irrelevant to the problem because 5210 is not mms capable. And neither n7650 nor t68i understand MMSURL in settings Yes, they do. We are provisioning MMSC OTA to Ericsson T68 and every Nokia that supports MMS (3510, 7650, 7210, etc). I can send some examples if required. -- Bruno Rodrigues [EMAIL PROTECTED] signature.asc Description: This is a digitally signed message part
Re: [RFC] unified-prefix on smsc basis
+1 for me - Original Message - From: Stipe Tolj [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Friday, September 06, 2002 10:58 AM Subject: [RFC] unified-prefix on smsc basis Hi all, I'd like to introduce a 'unified-prefix' directive on a smsc basis that behaves the same way as the global one defines in the core group. Main intention is that smsbox does route the msg to bearerbox and bearerbox may normalize the number according to the global unified-prefix rules. But running bearerbox with dozens of various smsc connections require to map the number in different formats, which is not possible using one global unified-prefix rule set. I.e. thing of smsbox sends a msg with destination 49173... (hence in international format). Now unified-prefix convers the number to 0049173... because it's required for one EMI link. If we want to route it to an other smsc which uses SMPP with dest-addr-ton = 1, then we'd except to pass it in 49173... format. So, I guess defining the unified-prefix rules on a smsc basis is very easy to implement. BTW, I'm just coding this right now. Any comments or objections against introducing this new feature?! Stipe [EMAIL PROTECTED] --- Wapme Systems AG Vogelsanger Weg 80 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
[PATCH] Still with lots of work...
...but I'm missing some good old C coding. I've commited (tested on linux and windows) a patch to enable including a directory full of configuration files. Just: include = configdir and kannel would detect it as a directory and include every file inside it. Besides, I've compiled a cygwin version, zipped and depoyed it on http://kannel.org/download/kannel-2002.09.07.zip Please comment it to my mail regards
Re: [patch] smsbox memory leaks and service header
On Sun, 2002-06-02 at 09:34, Stipe Tolj wrote: Angel Fradejas wrote: Hi all, here's a patch for smsbox.c 1) a couple of memory leaks (msg-sms.udhdata and accepted_chars) added the fixes to cvs, thanks Angel! 2) allow to change msg-sms.service via X-Kannel-Service header or xml tag I'm +0 for this. I don't see usage szenarios for these kind of support. Can somebody please enlight me if I'm missing the obvious. And if you wish to let the application differentiate the service it is using, just use the Account field. Stipe [EMAIL PROTECTED] --- Wapme Systems AG Münsterstr. 248 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
Re: [patch] smsbox memory leaks and service header
On Fri, 2002-05-31 at 17:08, Angel Fradejas wrote: Hi all, here's a patch for smsbox.c 1) a couple of memory leaks (msg-sms.udhdata and accepted_chars) thanks, could someone else apply this ? 2) allow to change msg-sms.service via X-Kannel-Service header or xml tag Why should you change the sms.service ? sms.service is the name field in configuration, for sms-service and sendsms-user, so you can identify the service or username used, in bearerbox_access.log. We shouldn't give access to clients connecting through http. I vote -1 for this part Cheers. Angel Fradejas Mediafusión España, S.A. [EMAIL PROTECTED] www.mediafusion.es Tel. +34 91 252 32 00 Fax +34 91 572 27 08
Re: EMI: Serious Problem PANIC: Too many concurrent allocations
On Sat, 2002-06-01 at 02:11, Jari Juslin wrote: Kalle Marjola wrote: On 24 May 2002, Abd Rahman Johari wrote: I'm trying to send out SMS Broadcast and after a while kannel just die because of PANIC: Too many concurrent allocations You need to compile Kannel with non-checking malloc, i.e. use native malloc. ./configure --with-defaults=speed or --with-malloc=native should help. Further question: In general situation, does this really correct problems or just hide them? I though that the checking malloc spots errors and making error messages to go away by just not checking anymore sounds quite hazard.. Or is it so, that this Too many concurrent allocations checking failing does not always mean there is an error? I have got the smsbox to panic this way when put under too heavy load and bearerbox when restoring big enough store file (1000+ messages), and these sound like this message would occur more often when Kannel is put under high pressure (and possibly making much concurrent allocations for some good reason?). The concurrent allocations are relative to the checking malloc, that is limited (or configured somewhere in the code) to 1 or 2 MB. If you use native malloc, the limit should be the memory available. I've sent 200K (10k at once, with post-xml) messages and had the store.lock as bigger than 40MB, with 30k messages in queue. I don't recall now the exact values, but I were always expecting to see kannel crashing somewhere, but it worked out ok.
RE: Daily patch: gateway
On Fri, 2002-05-31 at 11:49, Harrie Hazewinkel wrote: --On Friday, May 31, 2002 12:07 PM +0200 Angel Fradejas [EMAIL PROTECTED] wrote: Harrie please, Would you be so kind as to comment these changes in ChangeLog? Done. They're somewhat described in a post from May 25 (subject: changes to the heartbeat code.) The message is the same as the commit message. I personally, would prefer if the CVS-message during commit is used for this. From our former architecture, Lars: (I hope everything is ok with ex-wapit people) ChangeLog We use the file called `ChangeLog' to log changes, not CVS commit entries. CVS entries are not usable by people who don't have CVS access (e.g., they are off-line at the moment, which happens often for laptop users), and are also harder to browse than a single file. When writing an entry, please try to make sure the entry is at least somewhat understandable without referring to the actual patch it reflects. This helps immensely those who want to follow Kannel development without having to wade through millions of lines of patches per year. Harrie Internet Management Consulting mailto:[EMAIL PROTECTED]http ://www.mod-snmp.com/
Re: Patch submission and release policy (Was: [PATCH] problems withHTTPS and base support for per message billing)
On Wed, 2002-05-29 at 08:39, Harrie Hazewinkel wrote: --On Wednesday, May 29, 2002 3:45 AM +0100 Bruno Rodrigues [EMAIL PROTECTED] wrote: On Mon, 2002-05-20 at 15:16, Stipe Tolj wrote: Oded Arbel wrote: Agreed. I was hoping that at least the billing issue (I remember it being talked about in the list a while back) would interest people. I do think, though, that fixes to problems not yet detected in the wild should go in anyway : that's why it's called a development tree, if the solution does not break anything - of course. IMHO, the current situation where the CVS build must never be broken because it is the production version and so patches require careful scrutiny before going in is not healthy. CVS _is_ the place to test fixes and new features - when you require that people will download and apply your patches one by one, the number of testers will shrink to the number of people interested in the specfic patch - which in a not-so-high visibility project like Kannel could easily get down to 1~2 people - or even less. I agree. CVS is for development and users do not like to apply many patches themselves in order to get it into a certain state. But kannel have some special kind of users. Besides the ones that wants to test kannel against a modem or mobile, the majority are almost power users that don't get scared with terminology of stable and unstable. It's like beeing afraid of debian unstable :P I'm always using cvs in production. Some bugs are only visible on production systems and I don't have time to do testings before upgrading. And if some message is lost, I can always blame the SMSC ;) You must be a lucky chap then. :-) I just have the facility to being able to have a production system to do debugging, and some knowledge to fix quickly if something goes wrong (and there's always the local bearerbox.old and smsbox.old ;) )
Re: Patch submission and release policy (Was: [PATCH] problems with HTTPS and base support for per message billing)
On Mon, 2002-05-20 at 15:16, Stipe Tolj wrote: Oded Arbel wrote: Agreed. I was hoping that at least the billing issue (I remember it being talked about in the list a while back) would interest people. I do think, though, that fixes to problems not yet detected in the wild should go in anyway : that's why it's called a development tree, if the solution does not break anything - of course. IMHO, the current situation where the CVS build must never be broken because it is the production version and so patches require careful scrutiny before going in is not healthy. CVS _is_ the place to test fixes and new features - when you require that people will download and apply your patches one by one, the number of testers will shrink to the number of people interested in the specfic patch - which in a not-so-high visibility project like Kannel could easily get down to 1~2 people - or even less. case in point is the +CMTI patch by Alex Judd - it seems like a perfectly valid feature, but only 2 or 3 people on this list are at the same time interested and skilled to test iX-Mozilla-Status: 0009tences where some of them cannot find the time to do so, this perfectly good feature would simply be abandoned. I suggest we should roll out a release ASAP, using the following schedule : - branch the tree now (yesterday would have been a good time too ;-) and label it 1.2.0. - bug fixes may be submitted to either of the trees, and then ported to the other. - new features may be submitted only to the HEAD tree. - features and bug fixes will be submitted freely to the HEAD tree with minimum checks for style and obvious coding errors. - the HEAD tree will be considered unstable and fit only for development work. Using this method we would not further degrade the current situation (where people who have problems are told to upgrade their production servers to the CVS version - as it is more stable), while stabilizing the development effort for a full fledged stable release w/o hampering further feature development. Opinions please ? +1 for most of that. I was anyway concidering asking the developers about releasing 1.2.0. I'd like to hear from Bruno, Andreas and some others what they think about if current CVS HEAD is stable enough to make it a stable release 1.2.0? Well.. To be honest, using the CVS is an advantage because that way we get 100% testing and debug, code is done with less errors and bugs are fixed quicker ;) I'm always using cvs in production. Some bugs are only visible on production systems and I don't have time to do testings before upgrading. And if some message is lost, I can always blame the SMSC ;) There's some structural changes that we should do, and for that we really need a different branch. Modularity, new autoconf, real unicode support, etc. But for that, before thinking in branches and releases, we should think in the new architecture. Stipe [EMAIL PROTECTED] --- Wapme Systems AG Münsterstr. 248 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
Snapshots aren't being built (Was: Re: smsc_at2.c at+cmms=2)
- Original Message - From: Andrea Viscovich [EMAIL PROTECTED] To: Alex Judd [EMAIL PROTECTED]; Oded Arbel [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Thursday, April 18, 2002 3:51 PM Subject: Re: smsc_at2.c at+cmms=2 Thanks Alex and Oded. but latest CVS has a configuration for that (in modems.conf) that is disabled by default for most modem types I just dowloaded latest tar.gz (snapshot). There is exactly the same code: http://www.kannel.3glab.org/cgi-bin/viewcvs.cgi/gateway/gw/smsc_at2.c?annota te=1.44#1425 Snapshots aren't beeing built since Apr 09, and I've commited in Apr 10. Please try the cvs version for now
Re: [Fwd: Daily patch: gateway]
- Original Message - From: Stipe Tolj [EMAIL PROTECTED] To: David Holland [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Thursday, April 18, 2002 4:30 PM Subject: Re: [Fwd: Daily patch: gateway] David Holland wrote: On Thu, Apr 18, 2002 at 03:29:42PM +0100, Bruno David Simões Rodrigues wrote: Hei, what happened to changelog ?!?!?! my fault, sorry. Cygwin's DOS/UNIX line endings got messed up in WinCVS, so that's why there way a big diff, actually containing the same. That's why I always do a cvs diff and recheck the patch agains the changelog before commiting :P Stipe [EMAIL PROTECTED] --- Wapme Systems AG Münsterstr. 248 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
Re: [Fwd: Daily patch: gateway]
- Original Message - From: Stipe Tolj [EMAIL PROTECTED] To: Bruno Rodrigues [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Thursday, April 18, 2002 4:31 PM Subject: Re: [Fwd: Daily patch: gateway] Bruno Rodrigues schrieb: Can't we split changelog to some kind of old changelog.1.1.0, changelog.1.1.1 to prevent it from beeing so large ? Good idea. Do you intend to branch it off or are we going to cut it off any continue in the HEAD branch?! I was thinking in just splitting it, but I don't know where to. doc/old/changelog.*.*.* ? Stipe [EMAIL PROTECTED] --- Wapme Systems AG Münsterstr. 248 40470 Düsseldorf Tel: +49-211-74845-0 Fax: +49-211-74845-299 E-Mail: [EMAIL PROTECTED] Internet: http://www.wapme-systems.de --- wapme.net - wherever you are
Re: Coreaccess Log Parser
I have a set of perl scripts to split and convert from access.log to CLF and process it through webalizer As I fix some last minor issues, I'll commit them to contrib directory - Original Message - From: Cipher Strength [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Thursday, April 18, 2002 7:36 PM Subject: Coreaccess Log Parser Hi All, Is there any Coreaccess Log parser. If some body had make it, can share it with me. Currently i am using GREP unix command to parse log file or by inserting the Log file in DATABASE. Also if i want to make a pull service but not use any keyword neither default service any soloution?(might look funny). Also in every pull mobile can be unique can we get some thing more unique example if messages are sent to mobile from two web user and mobile user reply to it, how to distinguish to whom web user the message should be sent. Which version to use between these two Daily CVS or Development Release 1.1.6 Regards CIPHER _ MSN Photos is the easiest way to share and print your photos: http://photos.msn.com/support/worldwide.aspx
Re: Problems with Portuguese
That's what I meant. The page doesn't have an encoding, so it defaults to UTF8, and bruno@server:~$ GET -H 'Accept: text/vnd.wap.wml' 'wap.copel.com' | xmllint - -:9: error: Input is not proper UTF-8, indicate encoding ! a href=consultapro.wmlAndamento serviços/a ^ -:9: error: Bytes: 0xE7 0x6F 0x73 0x3C a href=consultapro.wmlAndamento serviços/a ^ I've tested with CMG WapGW and it works (I get the page) People, may we have a default encoding if wml doesn't have one ? or perhaps a list, so kannel could try to encode it with utf8 and then with iso-8859-1 or something ? - Original Message - From: Juvinski [EMAIL PROTECTED] To: Bruno Rodrigues [EMAIL PROTECTED] Sent: Thursday, April 18, 2002 9:22 PM Subject: Re: Problems with Portuguese Motorola v60, nokia 3320, 7110, Ericsson R300 and t60. What happened is that I don't have access to the page(www.copel.com) so I can't modify the the page. But where I need modify this parameters ??? Thank 's Juvinski Bruno Rodrigues wrote: Uh ? I'm serving wap pages with accented chars perfectly. Do you have the encoding in wml ? ?xml version=1.0 encoding=iso-8859-1? I'm using -1 and -15, even if I still can't set the EUR sign :( (BTW: my CMG Wap gw doesn't understand iso-8859-15) Which mobile is it ? - Original Message - From: Juvinski [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Thursday, April 18, 2002 7:36 PM Subject: Problems with Portuguese Hi, I'm testing the versions 1.1.5 and 1.1.6 (WAP), and there are some pages with the characters like ã, ç, õ and some others. And the mobile device freeze the wapbox result in a segment fault. Someone know how I can resolv this problem. Thank's Vinicius Juvinski
[RFC] cgi value = 0 behaviour
To enable smsbox to accept '0' values in cgi fields, the interface should change and the behaviour will be broken, but I think this should be done before we launch the new stable version (1.2). May I change the code and documentation now ? What will be changed is : mclass: instead of 0-4, will be 0-3 and -1 or not present for default pid: 0 really means 0 and -1 or not present means not present in pdu coding: 0-3 means 7bit, 8bit or ucs2 and -1 or not present means default behaviour of kannel defining it with udh presence compress: 0,1 forces no-compression (or decompression) or compression, and -1 means whatever (I haven't yet done the code to compression) validity: 0 means error (kannel can't send a message that already expired) and -1/not defined means smsc defaults deferred: 0 means now, -1 or not defined means now too ;) (or when kannel wants to send it?) alt-dcs: 0 means 0x0*, 1 means 0xF*, -1 means let kannel do what it wants So, it's only cgi fields introduced in 1.1.5++ versions. What do you think ?
[BUG] Alternate EMI2 SMSC
There's still a bug in emi2 that I'm now hunting: Context: SMSC connection is closed. (retry config value = true). Kannel tries to reconnect to primary smsc. Everything works ok If kannel cannot connect to primary and can connect to alternate, it won't send any message, not even the forced keepalive. Something related to keepalive_time or idletimeout_time. I'll commit a patch as soon as I find the bug
Re: Disabling the auto response to reply from get-url
max-messages = 0 I think... confirm on the documentation - Original Message - From: Alex Judd [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Thursday, April 04, 2002 5:13 PM Subject: Disabling the auto response to reply from get-url Probably an obvious question but how do I stop the automatic SMS back to the user of any response triggered by the get-utl I specify. For example get-url = http://do something and when I run do-something it replies 'done', however I don't want to send 'done' to the user. ? Thanks Alex -- Alex Judd http://www.skywire.co.uk
