Re: [Kannel-Users] Re: [Kannel-Users] Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-08 Thread Peter Beckman
On Sat, 7 Aug 2004, Alexander Malysh wrote:

  On Thu, 2004-08-05 at 23:54, Peter Beckman wrote:
 
  In 1.3.1 I'm seeing that when the remote connection disconnects without
  an unbind, and then reconnects, messages stop getting sent until either
  one side or I restart the gateway.
 
  Anyone else see this?

 which smsc module? more details with debug logs please...

 SMSC SMPP module.  The problem is transient (or at least seemingly so).
 The remote SMPP server sends an unbind and disconnects.  Kannel attempts
 to reconnect every 10 seconds.  Once the remote SMPP server opens for
 business again, Kannel can connect (or at least it looks that way), but
 does not send messages.  A restart on either end will begin sending the
 messages again.

 Ok, here's what happens:

 The last message is successfully sent to smpphost

  2004-07-16 21:30:00 [7] DEBUG: Looking for DLR smsc=smpphost, ts=17869084, 
dst=12125551213, type=12

 A few seconds later the remote SMPP host disconnects, sending an unbind
 for both TX and RX.

 2004-07-16 21:30:14 [8] DEBUG: SMPP PDU 0xac803c90 dump:
 2004-07-16 21:30:14 [8] DEBUG:   type_name: unbind
 2004-07-16 21:30:14 [8] DEBUG:   command_id: 6 = 0x0006
 2004-07-16 21:30:14 [8] DEBUG:   command_status: 0 = 0x
 2004-07-16 21:30:14 [8] DEBUG:   sequence_number: 624 = 0x0270
 2004-07-16 21:30:14 [8] DEBUG: SMPP PDU dump ends.
 2004-07-16 21:30:15 [8] ERROR: SMPP[smpphost]: I/O error or other error. 
Re-connecting.
 2004-07-16 21:30:15 [8] ERROR: connect failed
 2004-07-16 21:30:15 [8] ERROR: System error 111: Connection refused
 2004-07-16 21:30:15 [8] ERROR: error connecting to server `12.34.56.78' at port `8202'
 2004-07-16 21:30:15 [8] ERROR: SMPP[smpphost]: Couldn't connect to server.
 2004-07-16 21:30:15 [8] ERROR: SMPP[smpphost]: Couldn't connect to SMS center 
(retrying in 10 seconds).
 2004-07-16 21:30:19 [7] DEBUG: SMPP[smpphost]: Got PDU:
 2004-07-16 21:30:19 [7] DEBUG: SMPP PDU 0x873bfe0 dump:
 2004-07-16 21:30:19 [7] DEBUG:   type_name: unbind
 2004-07-16 21:30:19 [7] DEBUG:   command_id: 6 = 0x0006
 2004-07-16 21:30:19 [7] DEBUG:   command_status: 0 = 0x
 2004-07-16 21:30:19 [7] DEBUG:   sequence_number: 625 = 0x0271
 2004-07-16 21:30:19 [7] DEBUG: SMPP PDU dump ends.
 2004-07-16 21:30:34 [7] DEBUG: SMPP[smpphost]: Sending enquire link:
 2004-07-16 21:30:34 [7] DEBUG: SMPP PDU 0x873bfe0 dump:
 2004-07-16 21:30:34 [7] DEBUG:   type_name: enquire_link
 2004-07-16 21:30:34 [7] DEBUG:   command_id: 21 = 0x0015
 2004-07-16 21:30:34 [7] DEBUG:   command_status: 0 = 0x
 2004-07-16 21:30:34 [7] DEBUG:   sequence_number: 21266 = 0x5312
 2004-07-16 21:30:34 [7] DEBUG: SMPP PDU dump ends.
 2004-07-16 21:30:34 [7] DEBUG: SMPP[smpphost]: Got PDU:
 2004-07-16 21:30:34 [7] DEBUG: SMPP PDU 0x873bfe0 dump:
 2004-07-16 21:30:34 [7] DEBUG:   type_name: enquire_link_resp
 2004-07-16 21:30:34 [7] DEBUG:   command_id: 2147483669 = 0x8015
 2004-07-16 21:30:34 [7] DEBUG:   command_status: 0 = 0x
 2004-07-16 21:30:34 [7] DEBUG:   sequence_number: 21266 = 0x5312
 2004-07-16 21:30:34 [7] DEBUG: SMPP PDU dump ends.
 2004-07-16 21:30:34 [7] ERROR: SMPP[smpphost]: I/O error or other error. 
Re-connecting.
 2004-07-16 21:30:34 [7] ERROR: connect failed
 2004-07-16 21:30:34 [7] ERROR: System error 111: Connection refused
 2004-07-16 21:30:34 [7] ERROR: error connecting to server `12.34.56.78' at port `8203'
 2004-07-16 21:30:34 [7] ERROR: SMPP[smpphost]: Couldn't connect to server.
 2004-07-16 21:30:34 [7] ERROR: SMPP[smpphost]: Couldn't connect to SMS center 
(retrying in 10 seconds).

H.  I get an unbind for the TX session #7, yet the system is still
connected enough to do an enquire_link and get a response, then it dies.

So it stays offline for a little bit then comes back up.

 2004-07-16 21:30:44 [7] DEBUG: SMPP[smpphost]: Sending PDU:
 2004-07-16 21:30:44 [7] DEBUG: SMPP PDU 0x8744fe8 dump:
 2004-07-16 21:30:44 [7] DEBUG:   type_name: bind_transmitter
 2004-07-16 21:30:44 [7] DEBUG:   command_id: 2 = 0x0002
 2004-07-16 21:30:44 [7] DEBUG:   command_status: 0 = 0x
 2004-07-16 21:30:44 [7] DEBUG:   sequence_number: 21267 = 0x5313
 2004-07-16 21:30:44 [7] DEBUG:   system_id: x
 2004-07-16 21:30:44 [7] DEBUG:   password: y
 2004-07-16 21:30:44 [7] DEBUG:   system_type: VMA
 2004-07-16 21:30:44 [7] DEBUG:   interface_version: 52 = 0x0034
 2004-07-16 21:30:44 [7] DEBUG:   addr_ton: 0 = 0x
 2004-07-16 21:30:44 [7] DEBUG:   addr_npi: 0 = 0x
 2004-07-16 21:30:44 [7] DEBUG:   address_range: 
 2004-07-16 21:30:44 [7] DEBUG: SMPP PDU dump ends.
 2004-07-16 21:30:45 [7] DEBUG: SMPP[smpphost]: Got PDU:
 2004-07-16 21:30:45 [7] DEBUG: SMPP PDU 0x8744fe8 dump:
 2004-07-16 21:30:45 [7] DEBUG:   type_name: bind_transmitter_resp
 2004-07-16 21:30:45 [7] DEBUG:   command_id: 2147483650 = 0x8002
 2004-07-16 21:30:45 [7] DEBUG:   command_status: 0 = 0x
 2004-07-16 21:30:45 [7] DEBUG:   

Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-07 Thread Stipe Tolj
Pedro B. wrote:
Bruno said he had to do the pack before August 13th.
I swear i looked at the calendar just now and appeared to me that 1 week 
from now is past August 13th.
now, to make things more easy, we can roll it even earlier. I'll have 
to catch-up with the list and mantis, since I was sort of 
unavailable the last 1-2 weeks due to other issues.

When we are green for release, we will roll 1.4.0 stable then. So 
Bruno will be able to package it for Debian. ;)

@Bruno: is this a good offer?
Stipe
mailto:stolj_{at}_wapme.de
---
Wapme Systems AG
Vogelsanger Weg 80
40470 Düsseldorf, NRW, Germany
phone: +49.211.74845.0
fax: +49.211.74845.299
mailto:info_{at}_wapme-systems.de
http://www.wapme-systems.de/
---
-BEGIN PGP PUBLIC KEY BLOCK-
Version: GnuPG v1.2.2 (Cygwin)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=aYCI
-END PGP PUBLIC KEY BLOCK-


Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-07 Thread Bruno Rodrigues
Stipe Tolj wrote:
Alexander Malysh wrote:
Hello Bruno,
1.3.2 should be good tested and remaining bug's should be fixed first 
before
we can call it stable (IMO) and cvs version (before 1.3.2) was not tested
by too much people, so we should wait a bit...

P.S. why not package 1.3.2 for debian and get it more testing?
Bruno Rodrigues wrote:

Hello all.
Does anyone have any complain about 1.3.2 in comparison with 1.2.1 ?
What do you think about releasing 1.3.2 as 1.2.2 stable so I can package
it for Debian and we could have the latest stable version in Sarge ?
Stipe ?

I'd like to go for the 1.4 stable branch, rather then having this major 
step under 1.2.

This is historically also been made for 1.3 devel branch.
So, I'd like to see 1.4.0 stable rolled in about a week, to give people 
time to identify any remaining open issues and get the issues out of the 
BTS too.
All apps have bugs
I'm seeing two patches from 1.3.2 to HEAD, and they seem fine to me. I'm 
fixing now the double recode bug.

If you don't mind, I'll call it 1.4.0 then.
get back in a few minutes...





Re: [Kannel-Users] Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-07 Thread Alexander Malysh
Hi,

Sahil Aggarwal wrote:

 i have lost many messages because of this type of disconnection as
 everytime i have to restart the gateway.

if you don't use store-file then it's your problem...

 
 Sahil
 
 
 On Thu, 2004-08-05 at 23:54, Peter Beckman wrote:
 
 On Thu, 5 Aug 2004, Bruno Rodrigues wrote:
 
  Everybody else, please vote now or shutup forever ;) (how is it the
  setence when you marry, in english ?)
 
  Speak now or forever hold your peace.
 
 In 1.3.1 I'm seeing that when the remote connection disconnects without
 an unbind, and then reconnects, messages stop getting sent until either
 one side or I restart the gateway.
 
 Anyone else see this?

which smsc module? more details with debug logs please...

 
 Beckman

---
 Peter Beckman  Internet
 Guy
 [EMAIL PROTECTED]
 http://www.purplecow.com/

---
 


-- 
Thanks,
Alex




Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-07 Thread Stipe Tolj
Pedro B. wrote:
Can this aspell version replace the www docs too?
yep, I'll hit it towards the htdocs dir ;)
Stipe
mailto:stolj_{at}_wapme.de
---
Wapme Systems AG
Vogelsanger Weg 80
40470 Düsseldorf, NRW, Germany
phone: +49.211.74845.0
fax: +49.211.74845.299
mailto:info_{at}_wapme-systems.de
http://www.wapme-systems.de/
---
-BEGIN PGP PUBLIC KEY BLOCK-
Version: GnuPG v1.2.2 (Cygwin)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=aYCI
-END PGP PUBLIC KEY BLOCK-


Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-06 Thread Bruno Rodrigues
Alan McNatty wrote:
I agree completely - just highlighting for Bruno. 

If he's going to make a Debian package for stable it will be an issue if
package has is compiled --with-pgsql and as long as 7.2 is in stable.  

On Fri, 2004-08-06 at 10:04, Pedro B. wrote:
Sorry, but doing patches to support a 7.2.x version when 7.4.x is the 
stable version, is like proposing to go back to stoneage and make fire 
by using 2 sticks.

The 7.2 is completely deprecated on a postgres point of view.
I do know debian-stable still uses it, but it shouldn't. 7.4.3 should 
have gone out of testing ages ago. 7.5 is halfway under development, 
because everyone in the postgres world (hate these tags, but they are 
sometimes needed) has implemented 7.4.

These are the 2 cents from a deep-into-postgres-point of view.
No,. I'm doing a package for testing, Sarge, just before it gets released.
Sarge is using 7.4 (http://packages.debian.org/testing/misc/postgresql), 
so I guess this is not an issue, right ?




Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-06 Thread Stipe Tolj
Alexander Malysh wrote:
Hello Bruno,
1.3.2 should be good tested and remaining bug's should be fixed first before
we can call it stable (IMO) and cvs version (before 1.3.2) was not tested
by too much people, so we should wait a bit...
P.S. why not package 1.3.2 for debian and get it more testing?
Bruno Rodrigues wrote:

Hello all.
Does anyone have any complain about 1.3.2 in comparison with 1.2.1 ?
What do you think about releasing 1.3.2 as 1.2.2 stable so I can package
it for Debian and we could have the latest stable version in Sarge ?
Stipe ?
I'd like to go for the 1.4 stable branch, rather then having this 
major step under 1.2.

This is historically also been made for 1.3 devel branch.
So, I'd like to see 1.4.0 stable rolled in about a week, to give 
people time to identify any remaining open issues and get the issues 
out of the BTS too.

Stipe
mailto:stolj_{at}_wapme.de
---
Wapme Systems AG
Vogelsanger Weg 80
40470 Düsseldorf, NRW, Germany
phone: +49.211.74845.0
fax: +49.211.74845.299
mailto:info_{at}_wapme-systems.de
http://www.wapme-systems.de/
---
-BEGIN PGP PUBLIC KEY BLOCK-
Version: GnuPG v1.2.2 (Cygwin)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=aYCI
-END PGP PUBLIC KEY BLOCK-


Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-06 Thread Pedro B.
Bruno said he had to do the pack before August 13th.
I swear i looked at the calendar just now and appeared to me that 1 week 
from now is past August 13th.

\\pb
Stipe Tolj wrote:
Alexander Malysh wrote:
Hello Bruno,
1.3.2 should be good tested and remaining bug's should be fixed first 
before
we can call it stable (IMO) and cvs version (before 1.3.2) was not tested
by too much people, so we should wait a bit...

P.S. why not package 1.3.2 for debian and get it more testing?
Bruno Rodrigues wrote:

Hello all.
Does anyone have any complain about 1.3.2 in comparison with 1.2.1 ?
What do you think about releasing 1.3.2 as 1.2.2 stable so I can package
it for Debian and we could have the latest stable version in Sarge ?
Stipe ?

I'd like to go for the 1.4 stable branch, rather then having this major 
step under 1.2.

This is historically also been made for 1.3 devel branch.
So, I'd like to see 1.4.0 stable rolled in about a week, to give people 
time to identify any remaining open issues and get the issues out of the 
BTS too.

Stipe
mailto:stolj_{at}_wapme.de
---
Wapme Systems AG
Vogelsanger Weg 80
40470 Düsseldorf, NRW, Germany
phone: +49.211.74845.0
fax: +49.211.74845.299
mailto:info_{at}_wapme-systems.de
http://www.wapme-systems.de/
---
-BEGIN PGP PUBLIC KEY BLOCK-
Version: GnuPG v1.2.2 (Cygwin)
mIsEP6mcYwEEAMDnUiUwrbb+xwTFWN6TxF2+XZu7/alwJMeCwMBRvXtPZqfjpPhS
OkBpU0F4TrVuugz1HINTSaJTYq10AzDQXp5NkyWgckqW79nPAWuOX0dicbJk+cN2
nM2TI4KaxUDe6u8hghNEnH/i2lXsUu9apnP/iixzV81VC2je3uc9hZpnAAYptEVT
dGlwZSBUb2xqIChUZWNobm9sb2d5IENlbnRlciAmIFJlc2VhcmNoIExhYikgPHRv
bGpAd2FwbWUtc3lzdGVtcy5kZT6ItAQTAQIAHgUCP6mcYwIbAwYLCQgHAwIDFQID
AxYCAQIeAQIXgAAKCRABV0w1BqPYRuSqA/wPzsQxao2YePENCtgRTrO86U6zg3sl
OcS6CJFI4FZP5h/xD3GRsNH1+MPSvZlomDdpFnr547DGz/Kq9MXuQwVvlVig5yWZ
K5dtKp1r5YLhxJQBhfirZbRFFnYmf19f18J8OoS28tuFVftDl1AIwJS3HLyBTv6H
g2HyLAEKQIp30Q==
=aYCI
-END PGP PUBLIC KEY BLOCK-




Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-05 Thread Bruno Rodrigues
Hello all.
Does anyone have any complain about 1.3.2 in comparison with 1.2.1 ?
What do you think about releasing 1.3.2 as 1.2.2 stable so I can package 
it for Debian and we could have the latest stable version in Sarge ?

Stipe ?



Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-05 Thread Rogelio Serrano
On 2004-08-05 17:38:31 +0800 Bruno Rodrigues 
[EMAIL PROTECTED] wrote:

Hello all.
Does anyone have any complain about 1.3.2 in comparison with 
1.2.1 ?

What do you think about releasing 1.3.2 as 1.2.2 stable so I 
can package it 
for Debian and we could have the latest stable version in 
Sarge ?

Stipe ?

1.3.2 has a problem converting utf16be to utf-8. It happens 
when utf16be is first converted to latin-1 and when that fails 
it converted to utf8. When i commented out the conversion to 
latin 1 it worked.




Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-05 Thread Bruno Rodrigues
 Bruno Rodrigues wrote:

Hello all.

Does anyone have any complain about 1.3.2 in comparison with 1.2.1 ?

What do you think about releasing 1.3.2 as 1.2.2 stable so I can package
it for Debian and we could have the latest stable version in Sarge ?

Stipe ?
Alexander Malysh wrote:
Hello Bruno,
1.3.2 should be good tested and remaining bug's should be fixed first before
we can call it stable (IMO) and cvs version (before 1.3.2) was not tested
by too much people, so we should wait a bit...
P.S. why not package 1.3.2 for debian and get it more testing?
Because I can't pack a development version for a stable Debian 
release, and I have to do it before August 13th. I'd prefer to have a 
more current version than to keep older 1.2.1 in to-be-released Sarge.

Rogelio, can you please send me/us a patch with what you've changed ?
Is there any other bug that would be a regression from 1.2.1 ?



Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-05 Thread Alan McNatty
Hi Bruno,

I've tested the pgsql dlr storage on 7.2.1 (debian stable) but found
that Kannel utilised a DELETE ... LIMIT 1 in dlr_pgsql_remove from
gw/dlr_pgsql.c which is not supported in 7.2.1. 

ERROR: PGSQL: DELETE FROM dlr WHERE smsc='foo' AND ts='243251116' LIMIT
1;
ERROR: PGSQL: ERROR:  parser: parse error at or near LIMIT
ERROR: PGSQL: DB update failed!

For postgres stable you may want to change from ...

DELETE FROM a WHERE b=3 LIMIT 1;

- to - 

DELETE FROM a WHERE ctid=(SELECT ctid FROM a WHERE b=3 LIMIT 1);

Can supply a patch if required.

Cheers,
Alan


On Fri, 2004-08-06 at 01:09, Bruno Rodrigues wrote:
 Alexander Malysh wrote:
 
  hmm, ok - understand... looking through bug reports, I don't see any
  showstopper to call 1.3.2 - 1.2.2 (anyway 1.3.2 is more stable as 1.2.1),
  so I'm +1.
 
 Everybody else, please vote now or shutup forever ;) (how is it the 
 setence when you marry, in english ?)
 
  btw. copyright in debian directory doesn't match kannel license...
 Indeed. I haven't had time to update any debian stuff for too much time. 
   I'll check everything and post it on my site and kannel before I try 
 to upload to Debian. Hopefully, still this evening.
 
 
 




Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-05 Thread Pedro B.
Sorry, but doing patches to support a 7.2.x version when 7.4.x is the 
stable version, is like proposing to go back to stoneage and make fire 
by using 2 sticks.

The 7.2 is completely deprecated on a postgres point of view.
I do know debian-stable still uses it, but it shouldn't. 7.4.3 should 
have gone out of testing ages ago. 7.5 is halfway under development, 
because everyone in the postgres world (hate these tags, but they are 
sometimes needed) has implemented 7.4.

These are the 2 cents from a deep-into-postgres-point of view.
\\pb

Alan McNatty wrote:
Hi Bruno,
I've tested the pgsql dlr storage on 7.2.1 (debian stable) but found
that Kannel utilised a DELETE ... LIMIT 1 in dlr_pgsql_remove from
gw/dlr_pgsql.c which is not supported in 7.2.1. 

ERROR: PGSQL: DELETE FROM dlr WHERE smsc='foo' AND ts='243251116' LIMIT
1;
ERROR: PGSQL: ERROR:  parser: parse error at or near LIMIT
ERROR: PGSQL: DB update failed!
For postgres stable you may want to change from ...
DELETE FROM a WHERE b=3 LIMIT 1;
- to - 

DELETE FROM a WHERE ctid=(SELECT ctid FROM a WHERE b=3 LIMIT 1);
Can supply a patch if required.
Cheers,
Alan
On Fri, 2004-08-06 at 01:09, Bruno Rodrigues wrote:
Alexander Malysh wrote:

hmm, ok - understand... looking through bug reports, I don't see any
showstopper to call 1.3.2 - 1.2.2 (anyway 1.3.2 is more stable as 1.2.1),
so I'm +1.
Everybody else, please vote now or shutup forever ;) (how is it the 
setence when you marry, in english ?)


btw. copyright in debian directory doesn't match kannel license...
Indeed. I haven't had time to update any debian stuff for too much time. 
 I'll check everything and post it on my site and kannel before I try 
to upload to Debian. Hopefully, still this evening.







Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-05 Thread Alan McNatty
I agree completely - just highlighting for Bruno. 

If he's going to make a Debian package for stable it will be an issue if
package has is compiled --with-pgsql and as long as 7.2 is in stable.  

On Fri, 2004-08-06 at 10:04, Pedro B. wrote:
 Sorry, but doing patches to support a 7.2.x version when 7.4.x is the 
 stable version, is like proposing to go back to stoneage and make fire 
 by using 2 sticks.
 
 The 7.2 is completely deprecated on a postgres point of view.
 
 I do know debian-stable still uses it, but it shouldn't. 7.4.3 should 
 have gone out of testing ages ago. 7.5 is halfway under development, 
 because everyone in the postgres world (hate these tags, but they are 
 sometimes needed) has implemented 7.4.
 
 These are the 2 cents from a deep-into-postgres-point of view.
 
 \\pb
 
 
 
 
 Alan McNatty wrote:
 
  Hi Bruno,
  
  I've tested the pgsql dlr storage on 7.2.1 (debian stable) but found
  that Kannel utilised a DELETE ... LIMIT 1 in dlr_pgsql_remove from
  gw/dlr_pgsql.c which is not supported in 7.2.1. 
  
  ERROR: PGSQL: DELETE FROM dlr WHERE smsc='foo' AND ts='243251116' LIMIT
  1;
  ERROR: PGSQL: ERROR:  parser: parse error at or near LIMIT
  ERROR: PGSQL: DB update failed!
  
  For postgres stable you may want to change from ...
  
  DELETE FROM a WHERE b=3 LIMIT 1;
  
  - to - 
  
  DELETE FROM a WHERE ctid=(SELECT ctid FROM a WHERE b=3 LIMIT 1);
  
  Can supply a patch if required.
  
  Cheers,
  Alan
  
  
  On Fri, 2004-08-06 at 01:09, Bruno Rodrigues wrote:
  
 Alexander Malysh wrote:
 
 
 hmm, ok - understand... looking through bug reports, I don't see any
 showstopper to call 1.3.2 - 1.2.2 (anyway 1.3.2 is more stable as 1.2.1),
 so I'm +1.
 
 Everybody else, please vote now or shutup forever ;) (how is it the 
 setence when you marry, in english ?)
 
 
 btw. copyright in debian directory doesn't match kannel license...
 
 Indeed. I haven't had time to update any debian stuff for too much time. 
   I'll check everything and post it on my site and kannel before I try 
 to upload to Debian. Hopefully, still this evening.
 
 
 
  
  
  
  
 
 




Re: [Kannel-Users] Re: Is 1.3.2 stable enough to call it 1.2.2 ?

2004-08-05 Thread Sahil Aggarwal




i have lost many messages because of this type of disconnection as everytime i have to restart the gateway.

Sahil


On Thu, 2004-08-05 at 23:54, Peter Beckman wrote:

On Thu, 5 Aug 2004, Bruno Rodrigues wrote:

 Everybody else, please vote now or shutup forever ;) (how is it the
 setence when you marry, in english ?)

 Speak now or forever hold your peace.

In 1.3.1 I'm seeing that when the remote connection disconnects without an
unbind, and then reconnects, messages stop getting sent until either one
side or I restart the gateway.

Anyone else see this?

Beckman
---
Peter Beckman  Internet Guy
[EMAIL PROTECTED] http://www.purplecow.com/
---