icqlib-users

[icqlib-users] Is there a manual for icqlib???

[Miguel C. <ren...@gm...>]

Hello list

I was looking for a related site for the icq protocol, visited an old site but 
it is not working anymore and found icqlib, it may be what I need to create a 
basic icq client, but didn't find documentation on how to use it.

In 2001 I programmed a basic client with v2 and v3 protocol support but have 
not the code and documentation anymore. Want to use icqlib and save time, but 
anybody can help me to find some docs?

Thanks!!

[icqlib-users] New Project Using icqlib

[Matt T. <mat...@gm...>]

Hi,
I'm working on writing a native ICQ client for SkyOS
(http://www.skyos.org/).  I've successfully ported icqlib and am
working on the layout of the frontend currently.

The last release of icqlib was almost 5 full years ago.  I assume that
it isn't being maintained anymore.

My question is this:
Since it obviously isn't being maintained, should I find a different
ICQ library to base my client off of?  And if so, which?

Thanks in advance,

Matt

[icqlib-users] win32 client help!

[Ratnakar P. <Rat...@PL...>]

Hi all,
    Need some help for usage of icqlib on win NT, for a simple icq client
(that only sends messages to any uin)i'm developing. This is the sequence of
events:
1.call WinSock init(WSAStartup)
2.icq_Init()
3.icq_UnsetProxy();
4.icq_Connect()
5.icq_Login()
6. icq_SendMessage()
7.icq_HandleServerResponse();
8. icq_Logout(&link); 
9. icq_Disconnect(&link); 
10. icq_Done(&link); 
11. WSACleanup();

The Client is able to login fine but the messages that I send dont seem to 
be going through to the mentioned uins. I tried both
ICQ_SEND_DIRECT,ICQ_SEND_THRUSERVER.

Any help would be appreciated. Also please reply to the email address
mentioned as I am 
not subscribed to the mailing list...

pai

 

[icqlib-users] win32 Client

[Ratnakar P. <Rat...@PL...>]

Hi all,
    Need some help for usage of icqlib on win NT, for a simple icq client
(that only sends messages to any uin)i'm developing. This is the sequence of
events:
1.call WinSock init(WSAStartup)
2.icq_Init()
3.icq_UnsetProxy();
4.icq_Connect()
5.icq_Login()
6. icq_SendMessage()
7.icq_HandleServerResponse();
8. icq_Logout(&link); 
9. icq_Disconnect(&link); 
10. icq_Done(&link); 
11. WSACleanup();

The Client is able to login fine but the messages that I send dont seem to 
be going through to the mentioned uins. I tried both
ICQ_SEND_DIRECT,ICQ_SEND_THRUSERVER.

Any help would be appreciated. Also please reply to the email address
mentioned as I am 
not subscribed to the mailing list...

ratnakar

 

[icqlib-users] How not to write a client for icqlib.

[David S. <fes...@bt...>]

Just writing to say  patched together a small CLI client using icqlib.
It does all the things you'd expect it to do find it at, 
	http://dave.buhrer.net/gimbicq/ ,
and if you find it useful in some way then please let me know.

RE: [icqlib-users] Timeout and Logging in.

[Peter M. L. <pe...@le...>]

<Grin>

Well, when I started writing picq, I was just a student who didn't know
anything about open source software and application development. Then I got
a job and writing picq became even less oriented to writing something for
others. Maybe I should look into making a bit more documentation and
submitting it to freshmeat or something. :-)

Regards,

Peter.

-----Original Message-----
From: icq...@li...
[mailto:icq...@li...]On Behalf Of Denis V.
Dmitrienko
Sent: donderdag 7 juni 2001 21:33
To: icq...@li...
Subject: Re: [icqlib-users] Timeout and Logging in.


"Peter M. Lemmen" <pe...@le...> writes:

> As to the 'picq' name being taken... Hmm, well, I started on my client
back
> in the spring of 1999 (Heh, if you look at the epad and opad files, I've
not
> even touched them since november 1999...) and the first web page was up
even
> back then. So I'm not sure why I should even bother changing it. But if I
> come up with a better name (or anyone else has a good suggestion) I'll
think
> about it. :-)

There was no PICQ entries on the search engines back to the fall of
2000... ;) And right now search request on this word for example on Google
returns DataNaut's PICQ pages and not yours. But this is ok and you
shouldn't really worry about it since first of all, your PICQ is for Unix
platform and DataNaut's one is for Palm; and second, PICQ was a code name
for that project and supposed to be changed.

--
Denis V. Dmitrienko | E-mail: de...@de... | ICQ#: 5538614
Home page: http://denix.org | de...@kd...
History became legend... Legend became myth.
And some things that should not have been forgotten ...were lost.

_______________________________________________
icqlib-users mailing list
icq...@li...
http://lists.sourceforge.net/lists/listinfo/icqlib-users

Re: [icqlib-users] Timeout and Logging in.

[Denis V. D. <de...@nu...>]

"Peter M. Lemmen" <pe...@le...> writes:

> As to the 'picq' name being taken... Hmm, well, I started on my client back
> in the spring of 1999 (Heh, if you look at the epad and opad files, I've not
> even touched them since november 1999...) and the first web page was up even
> back then. So I'm not sure why I should even bother changing it. But if I
> come up with a better name (or anyone else has a good suggestion) I'll think
> about it. :-)

There was no PICQ entries on the search engines back to the fall of
2000... ;) And right now search request on this word for example on Google
returns DataNaut's PICQ pages and not yours. But this is ok and you
shouldn't really worry about it since first of all, your PICQ is for Unix
platform and DataNaut's one is for Palm; and second, PICQ was a code name
for that project and supposed to be changed.

-- 
Denis V. Dmitrienko | E-mail: de...@de... | ICQ#: 5538614
Home page: http://denix.org | de...@kd...
History became legend... Legend became myth.
And some things that should not have been forgotten ...were lost.

RE: [icqlib-users] Timeout and Logging in.

[Peter M. L. <pe...@le...>]

Hello,

Ok, thanks for the advice on icq_Login. I've been changing around the
synchronising of the contact list and I'll try adding the contacts before
logging in.

As to the 'picq' name being taken... Hmm, well, I started on my client back
in the spring of 1999 (Heh, if you look at the epad and opad files, I've not
even touched them since november 1999...) and the first web page was up even
back then. So I'm not sure why I should even bother changing it. But if I
come up with a better name (or anyone else has a good suggestion) I'll think
about it. :-)

Regards,

Peter.

-----Original Message-----
From: icq...@li...
[mailto:icq...@li...]On Behalf Of Denis V.
Dmitrienko
Sent: donderdag 7 juni 2001 20:07
To: icq...@li...
Cc: icqlib-dev mailing list
Subject: Re: [icqlib-users] Timeout and Logging in.


"Peter M. Lemmen" <le...@ho...> writes:

> I've been having an odd problem since using Icqlib 1.0.0 in my icq client.
> I've added support for handling Timeouts which works on the following
logic:
>
> When a icq_SetTimeout event is generated...
>   If the interval is -1, the current Timeout is cleared.
>   If the interval is 0, the current Timeout is cleared.
>   If the interval is larger then 0, the current Timeout is changed to
Now()
> + interval.
>
> When the Timeout passes...
>   icq_HandleTimeout is called.
>   And the Timeout is cleared.

Everything is fine so far.

> My login sequence is as follows:
>
> icq_Init is called to initialize the link.
> Callbacks are registered with the link.
> icq_Connect is called to create the UDP connection.
> icq_Login is called to send the login information to the server.
> The icq contacts are added to the link.
> icq_SendContactList is called.
> icq_SendVisibleList is called.

And this is wrong.
Move "The icq contacts are added to the link" step before "icq_Login is
called to send the login information to the server" because icq_Login()
sends all your contact information itself. It means you send your contact
list twice, that's why you experience timeouts.
In other words, you shouldn't call icq_SendContactList(),
icq_SendVisibleList() & icq_SendInvisibleList() functions manually on
login. Just populate those lists before calling icq_Login().
If you change your contact list after login (add new buddy), call
icq_SendNewUser() function (subject to rename!!!). Call
icq_SendVisibleList() & icq_SendInvisibleList() functions on any changes in
those lists as usual - there is no special functions to send single new
user added to them.

> I believe the icq_Login and/or the contact list calls generate a (15
second)
> Timeout event. Within those 15 seconds the user is logged in and all the
> UserOnline events are received.
> The timeout is not cleared however, and when the 15 seconds are up,
> HandleTimeout is called. A few seconds after that we get the whole list of
> UserOnline messages again, as if the contact list packets were resent.

It happens because ICQ server has special "feature" to drop packets when
they come too fast from the client, i.e. to prevent flood. Login procedure
generates a lot of the packets and some of them could be lost, that's why
we need Timeout manager - to resend unacknowledged (potentially lost)
packets.

> I think the Ack messages for the Contact list and Visible List get lost
> somewhere so icqlib is unable to match them up with the original sent
> packets.

Or just there was no Ack packets from the server at all...

> Is my handling logic of the Timeouts correct? Especially the double-clear
> case with a zero interval.

Yes, that is correct.

> Why is the contact list not acknowledged properly? Would it help if I
moved
> the sending of the contact and visible lists into the icq_Logged event?
(Ie,
> after the login was properly handled.)

Try to implement it like I said before.

> ----------------------------------------------------------------
> Peter M. Lemmen                               Peter's ICQ Client
> www.outreach.nl                             www.outreach.nl/picq
> pe...@le...                                pi...@ou...
> ----------------------------------------------------------------

As I already mentioned on the icqlib forum picq name is already taken:
http://www.datanaut.com/picq

--
Denis V. Dmitrienko | E-mail: de...@de... | ICQ#: 5538614
Home page: http://denix.org | de...@kd...
History became legend... Legend became myth.
And some things that should not have been forgotten ...were lost.

_______________________________________________
icqlib-users mailing list
icq...@li...
http://lists.sourceforge.net/lists/listinfo/icqlib-users

Re: [icqlib-users] Timeout and Logging in.

[Denis V. D. <de...@nu...>]

"Peter M. Lemmen" <le...@ho...> writes:

> I've been having an odd problem since using Icqlib 1.0.0 in my icq client.
> I've added support for handling Timeouts which works on the following logic:
> 
> When a icq_SetTimeout event is generated...
>   If the interval is -1, the current Timeout is cleared.
>   If the interval is 0, the current Timeout is cleared.
>   If the interval is larger then 0, the current Timeout is changed to Now()
> + interval.
> 
> When the Timeout passes...
>   icq_HandleTimeout is called.
>   And the Timeout is cleared.

Everything is fine so far.

> My login sequence is as follows:
> 
> icq_Init is called to initialize the link.
> Callbacks are registered with the link.
> icq_Connect is called to create the UDP connection.
> icq_Login is called to send the login information to the server.
> The icq contacts are added to the link.
> icq_SendContactList is called.
> icq_SendVisibleList is called.

And this is wrong.
Move "The icq contacts are added to the link" step before "icq_Login is
called to send the login information to the server" because icq_Login()
sends all your contact information itself. It means you send your contact
list twice, that's why you experience timeouts.
In other words, you shouldn't call icq_SendContactList(),
icq_SendVisibleList() & icq_SendInvisibleList() functions manually on
login. Just populate those lists before calling icq_Login().
If you change your contact list after login (add new buddy), call
icq_SendNewUser() function (subject to rename!!!). Call
icq_SendVisibleList() & icq_SendInvisibleList() functions on any changes in
those lists as usual - there is no special functions to send single new
user added to them.

> I believe the icq_Login and/or the contact list calls generate a (15 second)
> Timeout event. Within those 15 seconds the user is logged in and all the
> UserOnline events are received.
> The timeout is not cleared however, and when the 15 seconds are up,
> HandleTimeout is called. A few seconds after that we get the whole list of
> UserOnline messages again, as if the contact list packets were resent.

It happens because ICQ server has special "feature" to drop packets when
they come too fast from the client, i.e. to prevent flood. Login procedure
generates a lot of the packets and some of them could be lost, that's why
we need Timeout manager - to resend unacknowledged (potentially lost)
packets.

> I think the Ack messages for the Contact list and Visible List get lost
> somewhere so icqlib is unable to match them up with the original sent
> packets.

Or just there was no Ack packets from the server at all...

> Is my handling logic of the Timeouts correct? Especially the double-clear
> case with a zero interval.

Yes, that is correct.

> Why is the contact list not acknowledged properly? Would it help if I moved
> the sending of the contact and visible lists into the icq_Logged event? (Ie,
> after the login was properly handled.)

Try to implement it like I said before.

> ----------------------------------------------------------------
> Peter M. Lemmen                               Peter's ICQ Client
> www.outreach.nl                             www.outreach.nl/picq
> pe...@le...                                pi...@ou...
> ----------------------------------------------------------------

As I already mentioned on the icqlib forum picq name is already taken:
http://www.datanaut.com/picq

-- 
Denis V. Dmitrienko | E-mail: de...@de... | ICQ#: 5538614
Home page: http://denix.org | de...@kd...
History became legend... Legend became myth.
And some things that should not have been forgotten ...were lost.

[icqlib-users] Segmentation Faults on TCP Activity using ICQLib 1.0.0

[Peter M. L. <pe...@le...>]

Hello everyone,

I've been working on adding support for the TCP part of ICQLib to my client.
(Most notably, File transfers. :-) However I keep getting Segmentation
faults whenever any of the TCP stuff gets used. From sending direct
messages, to just receiving a file transfer request (and immediately
refusing it).

In the screenshot below I've tried to send a file from a windows client
(99b) to my client. It accepts the tcp connection and receives the file
request. Internally the file request is refused and the refusal message is
displayed properly by the windows client, so that works. However, about a
minute later, icqlib says it receives an unknown packet command '2050', and
again one minute later the client crashes on a segmentation fault.
(Note, there's two sets of Keepalive debug messages in there that are
unrelated I think, each from Timeout updated to Timeout stopped.)

Changing debug level to 10.
18:31:00 Debug[4] Icqlib: accepting tcp connection from 192.168.1.17:1160
18:31:01 Debug[4] Icqlib: hello packet received from 119893112 { version=6 }
18:31:01 rghost Received File Request.
Filename   : ALInvCont.wav
Description: Testing, 1, 2, 3.
18:31:23 Debug[3] Timeout updated, Now() + 15.
18:31:23 Debug[4] Icqlib: Send Keep Alive packet to the server
18:31:23 Debug[2] Sending Keepalive.
18:31:23 Debug[4] Icqlib: The server acknowledged the command
18:31:23 Debug[2] Server Acknowledged sequence 20650.
18:31:23 Debug[3] Timeout stopped on -1.
18:31:50 Debug[3] Icqlib: unknown packet command 2050!
18:32:23 Debug[3] Timeout updated, Now() + 15.
18:32:23 Debug[4] Icqlib: Send Keep Alive packet to the server
18:32:23 Debug[2] Sending Keepalive.
18:32:23 Debug[4] Icqlib: The server acknowledged the command
18:32:23 Debug[2] Server Acknowledged sequence 20651.
18:32:23 Debug[3] Timeout stopped on -1.
18:32:54        Awaiting user Command ============================ Onl ==
000 ==
./rt: line 3: 29112 Segmentation fault      ./picq ~/.picqrc-rts1 2>>rt.out

peter@ghost:~/dev/picq/current$


The same thing happens when I try to send a Message or Url direct, or
anything else related to TCP happens. The icq_TCPMain handler is called
about 4 times per second from within the main loop and I don't think I do
anything particularly weird with it.

Is there, apart from calling icq_TCPMain and registering the callbacks,
anything that needs to be done to initialize TCP handling properly?

Regards,

Peter.

------------------------------------------------------------
Peter M. Lemmen                           Peter's ICQ Client
www.outreach.nl                         www.outreach.nl/picq
pe...@le...                            pi...@ou...
------------------------------------------------------------

[icqlib-users] Timeout and Logging in.

[Peter M. L. <le...@ho...>]

Hello,

I've been having an odd problem since using Icqlib 1.0.0 in my icq client.
I've added support for handling Timeouts which works on the following logic:

When a icq_SetTimeout event is generated...
  If the interval is -1, the current Timeout is cleared.
  If the interval is 0, the current Timeout is cleared.
  If the interval is larger then 0, the current Timeout is changed to Now()
+ interval.

When the Timeout passes...
  icq_HandleTimeout is called.
  And the Timeout is cleared.


This seems to work pretty well, and I've not noticed a single dropped
message or status change since. However, it produces a slightly weird effect
on login.


My login sequence is as follows:

icq_Init is called to initialize the link.
Callbacks are registered with the link.
icq_Connect is called to create the UDP connection.
icq_Login is called to send the login information to the server.
The icq contacts are added to the link.
icq_SendContactList is called.
icq_SendVisibleList is called.

Then the program enters the main loop and starts handling server (and user)
response.

I believe the icq_Login and/or the contact list calls generate a (15 second)
Timeout event. Within those 15 seconds the user is logged in and all the
UserOnline events are received.
The timeout is not cleared however, and when the 15 seconds are up,
HandleTimeout is called. A few seconds after that we get the whole list of
UserOnline messages again, as if the contact list packets were resent.

I think the Ack messages for the Contact list and Visible List get lost
somewhere so icqlib is unable to match them up with the original sent
packets.

My questions:

Is my handling logic of the Timeouts correct? Especially the double-clear
case with a zero interval.

Why is the contact list not acknowledged properly? Would it help if I moved
the sending of the contact and visible lists into the icq_Logged event? (Ie,
after the login was properly handled.)


Regards,

Peter.

----------------------------------------------------------------
Peter M. Lemmen                               Peter's ICQ Client
www.outreach.nl                             www.outreach.nl/picq
pe...@le...                                pi...@ou...
----------------------------------------------------------------

[icqlib-users] Re: ICQLIB problems - solved!

[Ron P. <ic...@th...>]

Michael Hudson wrote:

> > So I've been attempting to get ICQLIB in my software.
> 
> Fair warning; I don't really know all that much about iclib (though I did
> get Python bindings written & got them to work).

NP.

> > It connects, then the hard drive gets accessed wildly.
> 
> Are you running out of memory (ie. is the hard disk access VM thrashing)?
> What OS?

Nah, my contact list counter (for sending the list) was overrunning the end
of the list.  So icqlib was storing loads of useless entries in its own
list and that was using up tonnes of swap space.  Heh heh.  FIXEd.

> AFAIK, icqlib doesn't touch the hard drive (except for file transfer
> stuff, obviously).

Yeah.  So far it's working nicely.  Have a temporary ICQ account created
for testing and debugging (don't know how to create accounts with icqlib
yet); logging in and out work, the userlist looks good.

> > 6) continuously call icq_Main in a loop, no delays
> 
> You don't really want to do that; you want to call icq_Main every 10 ms or
> so (try usleep or nanosleep).

Yeah, but that was a moot detail for the purposes at the time.  I've got
a usleep() working, but I'd rather use SIGALRM tuned to microseconds, and
delay for 50,000 of those.  (50ms).

> Could your callbacks be leaking memory?  That's about all I can think of.

Checking out the icqlib source, I noted that all callbacks are nullified,
so anything I don't use won't crash my program.  <Phew.>

> No, I think you can just implement the one's you're interested in; icqlib
> checks for NULLs before calling the callbacks.
> 
> HTH, but it might be better if someone with more iclib knowledge/intuition
> spoke up...

Your memory leak thing is always good to check out.  No one's written up
a decent icqlib tutorial, so I'm taking that task on.  Once the basics
are known, the whole development process is pretty simple.  Save the user
interface, of course.  Heh.  Still racking my brain about the best way to
do that under console.

(I'm thinking something like zicq would be nice, but have the userlist on
the left hand side of the screen and have scrolling and filtering caps.)

--
.--------------------------------------------.
|        Ron Poulton             Caffeinate. |
| EMail: thok at technologist.com      Code. |
|   Web: http://www.thok.net        Compute. |
`--------------------------------------------'

Zymurgy's Law of Volunteer Labor:
	People are always available for work in the past tense.

Re: [icqlib-users] Re: ICQLIB, problems.

[Ron P. <th...@te...>]

On Sun, 15 Apr 2001, Michael Hudson wrote:

> > So I've been attempting to get ICQLIB in my software.
> 
> Fair warning; I don't really know all that much about iclib (though I did
> get Python bindings written & got them to work).

NP.

> > It connects, then the hard drive gets accessed wildly.
> 
> Are you running out of memory (ie. is the hard disk access VM thrashing)?
> What OS?

Nah, my contact list counter (for sending the list) was overrunning the end
of the list.  So icqlib was storing loads of useless entries in its own
list and that was using up tonnes of swap space.  Heh heh.  FIXEd.

> AFAIK, icqlib doesn't touch the hard drive (except for file transfer
> stuff, obviously).

Yeah.  So far it's working nicely.  Have a temporary ICQ account created
for testing and debugging (don't know how to create accounts with icqlib
yet); logging in and out work, the userlist looks good.

> > 6) continuously call icq_Main in a loop, no delays
> 
> You don't really want to do that; you want to call icq_Main every 10 ms or
> so (try usleep or nanosleep).

Yeah, but that was a moot detail for the purposes at the time.  I've got
a usleep() working, but I'd rather use SIGALRM tuned to microseconds, and
delay for 50,000 of those.  (50ms).

> Could your callbacks be leaking memory?  That's about all I can think of.

Checking out the icqlib source, I noted that all callbacks are nullified,
so anything I don't use won't crash my program.  <Phew.>

> No, I think you can just implement the one's you're interested in; icqlib
> checks for NULLs before calling the callbacks.
> 
> HTH, but it might be better if someone with more iclib knowledge/intuition
> spoke up...

Your memory leak thing is always good to check out.  No one's written up
a decent icqlib tutorial, so I'm taking that task on.  Once the basics
are known, the whole development process is pretty simple.  Save the user
interface, of course.  Heh.  Still racking my brain about the best way to
do that under console.

(I'm thinking something like zicq would be nice, but have the userlist on
the left hand side of the screen and have scrolling and filtering caps.)

--
.--------------------------------------------.
|        Ron Poulton             Caffeinate. |
| EMail: thok at technologist.com      Code. |
|   Web: http://www.thok.net        Compute. |
`--------------------------------------------'

Zymurgy's Law of Volunteer Labor:
	People are always available for work in the past tense.

Re: [icqlib-users] Re: ICQLIB, problems.

[Michael H. <mw...@ca...>]

On Sat, 14 Apr 2001, Ron Poulton wrote:

> So I've been attempting to get ICQLIB in my software.

Fair warning; I don't really know all that much about iclib (though I did
get Python bindings written & got them to work).

> It connects, then the hard drive gets accessed wildly.

Are you running out of memory (ie. is the hard disk access VM thrashing)?
What OS?

AFAIK, icqlib doesn't touch the hard drive (except for file transfer
stuff, obviously).

> I did make a modification, but for contrast purposes, I was originally
> coding the ICQ prog to:
> 
> 1) icq_Init
> 2) set up callbacks
> 3) icq_UnsetProxy
> 4) icq_Connect
> 5) icq_Login if icq_Connect did not return -1
> 6) continuously call icq_Main in a loop, no delays

You don't really want to do that; you want to call icq_Main every 10 ms or
so (try usleep or nanosleep).

> Suepecting that icq_Main may have been getting overloaded, I modified the
> above structure such that:
> 
> 6) alarm(SIGALRM,bkgfuncs)
> 7) alarm(1)
> 8) the main loop would just check for a "Q" from the keyboard
> 
> The bkgfuncs, called once per second in theory, would call icq_Main and,
> after icq_Main was finished, rearm the system for a 1 second delay until
> the next call.
> 
> Now, the hard drive is still being accessed (it takes longer for it to
> become violent).

Could your callbacks be leaking memory?  That's about all I can think of.

> Is there something I'm missing in all of this?  Do I need to reference
> ALL callbacks, or is there a core set I need for just basic messaging
> and user databasing?

No, I think you can just implement the one's you're interested in; icqlib
checks for NULLs before calling the callbacks.

HTH, but it might be better if someone with more iclib knowledge/intuition
spoke up...

Cheers,
M.

[icqlib-users] Re: ICQLIB, problems.

[Ron P. <ic...@th...>]

So I've been attempting to get ICQLIB in my software.

It connects, then the hard drive gets accessed wildly.

I did make a modification, but for contrast purposes, I was originally
coding the ICQ prog to:

1) icq_Init
2) set up callbacks
3) icq_UnsetProxy
4) icq_Connect
5) icq_Login if icq_Connect did not return -1
6) continuously call icq_Main in a loop, no delays

Suepecting that icq_Main may have been getting overloaded, I modified the
above structure such that:

6) alarm(SIGALRM,bkgfuncs)
7) alarm(1)
8) the main loop would just check for a "Q" from the keyboard

The bkgfuncs, called once per second in theory, would call icq_Main and,
after icq_Main was finished, rearm the system for a 1 second delay until
the next call.

Now, the hard drive is still being accessed (it takes longer for it to
become violent).

Is there something I'm missing in all of this?  Do I need to reference
ALL callbacks, or is there a core set I need for just basic messaging
and user databasing?

--
.--------------------------------------------.
|        Ron Poulton             Caffeinate. |
| EMail: thok at technologist.com      Code. |
|   Web: http://www.thok.net        Compute. |
`--------------------------------------------'

Pushing 40 is exercise enough.

Re: [icqlib-users] NEWBIE ALERT.

[Ron P. <th...@te...>]

So I've been attempting to get ICQLIB in my software.

It connects, then the hard drive gets accessed wildly.

I did make a modification, but for contrast purposes, I was originally
coding the ICQ prog to:

1) icq_Init
2) set up callbacks
3) icq_UnsetProxy
4) icq_Connect
5) icq_Login if icq_Connect did not return -1
6) continuously call icq_Main in a loop, no delays

Suepecting that icq_Main may have been getting overloaded, I modified the
above structure such that:

6) alarm(SIGALRM,bkgfuncs)
7) alarm(1)
8) the main loop would just check for a "Q" from the keyboard

The bkgfuncs, called once per second in theory, would call icq_Main and,
after icq_Main was finished, rearm the system for a 1 second delay until
the next call.

Now, the hard drive is still being accessed (it takes longer for it to
become violent).

Is there something I'm missing in all of this?  Do I need to reference
ALL callbacks, or is there a core set I need for just basic messaging
and user databasing?

--
.--------------------------------------------.
|        Ron Poulton             Caffeinate. |
| EMail: thok at technologist.com      Code. |
|   Web: http://www.thok.net        Compute. |
`--------------------------------------------'

So, what's with this guy Gideon, anyway?  And why can't he ever
remember his Bible?

Re: [icqlib-users] NEWBIE ALERT.

[Ron P. <ic...@th...>]

On 3 Apr 2001, Denis V. Dmitrienko wrote:

> Ron Poulton writes:
> 
> > I just decided to check out icqlib.
> > 
> > I download zicq in order to gain some ground in utilizing the package, and
> > it looked good.  Connected, logged in just fine.
> 
> As I can see zicq has nothing to do with icqlib. So, everything you said is
> zicq problem and not icqlib, unless I missed something.

WHUPS, sorry about that.  "kicq" is what I was checking out.  "zicq" probably
came from some far reach of my synapses that have yet to be clear cut.

What I'm looking for is an example of how to login and logout and interpreting
callbacks - some basic rules and such.  My understanding so far is that the
icq_Init routine is called, the callbacks are all set up, the client calls
icq_Connect and icq_Login, interprets the callbacks, and either calls
icq_Logout or icq_Disconnect in order to remove itself from the ICQ network.

Anything I'm missing?

--
They spell it "da Vinci" and pronounce it "da Vinchy".  Foreigners
always spell better than they pronounce.
		-- Mark Twain

Re: [icqlib-users] NEWBIE ALERT.

[Denis V. D. <de...@nu...>]

ic...@th... writes:

> I just decided to check out icqlib.
> 
> I download zicq in order to gain some ground in utilizing the package, and
> it looked good.  Connected, logged in just fine.

As I can see zicq has nothing to do with icqlib. So, everything you said is
zicq problem and not icqlib, unless I missed something.

-- 
Denis V. Dmitrienko | E-mail: de...@de... | ICQ#: 5538614
Home page: http://denix.org | de...@kd...
History became legend... Legend became myth.
And some things that should not have been forgotten ...were lost.

[icqlib-users] NEWBIE ALERT.

[<ic...@th...>]

I just decided to check out icqlib.

I download zicq in order to gain some ground in utilizing the package, and
it looked good.  Connected, logged in just fine.

After a second, the program just stops and the hard drive goes wild.
The computer pauses (keyboard still works - LEDs toggle, can switch
from console to console), but everything else lags.  Even my MP3s
just stop playing.

This took an actual reset, as Ctrl-Alt-Del from terminal 1 wouldn't
even work.

Now, what am I missing?  This is almost as though something's recursing,
and eating up all the memory and process time.

Re: [icqlib-users] TCP problems.

[Bill S. <we...@ri...>]

On Thu, 1 Feb 2001, David B wrote:

> >2. Use the icq_Log callback to capture a log.  Log everything that
> >comes across the callback, try to send a TCP message to your program.
> >Send me the resulting output.
>
> I did this, and the log was working fine, logging everything.  However
> when sending a TCP message, nothing is logged, it just plain doesn't
> see it.  So I didn't bother attaching the log as it shows nothing about
> the tcp message or any errors about it.  Let me know if you still want
> it however.

Sure, send it along, I can at least look at it, maybe something will jump
out at me.

> >3. Give me a quick description of your system, running uname -a at a
> >shell prompt is fine.
>
> I probably should have mentioned this is a win32 machine and not a unix
> machine, and there no doubt lies the problem.

Aha!  Yes, this is important... we didn't do the win32 port ourselves, and
I know nothing about win32 development, so I'm afraid I won't be much help
:(

> In the headers I have, accept is defined as:
>
> SOCKET PASCAL accept(SOCKET,struct sockaddr *,int *);
>
> Hence why I changed socklen_t for int.  However I guess this wasn't a
> good thing to do.

Ah, actually you probably made the correct change.  I doubt that's the
reason accept isn't working.

> However I had the same tcp problem with icqlib1.0 and don't remember
> any socklen_t changing in there.  Or in fact changing anything.  I
> guess it's a winsock problem if it's working ok for you on a unix based
> system.

Most likely.  If you can debug the problem and provide a patch, we'd
gladly accept it.  If you have any questions about our code, I'll be more
than happy to help.

Bil

Re: [icqlib-users] TCP problems.

[David B <de...@so...>]

Sorry for the delay in replying.

>> Hi, I'm having a problem with the tcp part of icqlib.
>> 
>> I can do everything on the UDP side of things, receive msg's, send
>> messages (through the server) etc.  However if I enable tcp, messages
>> sent to me just time out and never appear.  However if I message the
>> person (ICQ_SEND_DIRECT) then I can get msg's from the person.
>> 
>> I have the socket handler stuff set up and it seems to be working ok.
>>  
>>  I also have the same problem if I call icq_Main on a 100ms timer.
>> 
>> Have I not set up something vital?  
>
>It sounds like you're doing everything correctly.  The socket handler
>callback is only there as an option - icqlib should work the same as it
>used to if you're using the timer method to call icq_Main, it's just 
>going through my new socket manager code.
>
>I've been getting a number of different reports for this lately, but 
>on my local system here, everything appears to work fine in both 
>directions.  If you could, please do the following:
>
>1. Ensure you've updated to the most recent CVS version - I put a few
>fixes in for the socket manager code after my initial checkin.

Yep it's the latest.

>
>2. Use the icq_Log callback to capture a log.  Log everything that 
>comes across the callback, try to send a TCP message to your program.  
>Send me the resulting output.

I did this, and the log was working fine, logging everything.  However 
when sending a TCP message, nothing is logged, it just plain doesn't 
see it.  So I didn't bother attaching the log as it shows nothing about 
the tcp message or any errors about it.  Let me know if you still want 
it however. 

>
>3. Give me a quick description of your system, running uname -a at a 
>shell prompt is fine.

I probably should have mentioned this is a win32 machine and not a unix 
machine, and there no doubt lies the problem. 

I do however remember something that is no doubt the cause of this (and 
therefore my fault).  When compiling icqlib (socketmanager.c to be 
precise) I got an error about no socklen_t.  I quickly checked the 
winsock headers and saw that the function being passes a socklen_t 
variable was 'int' in the headers and so I changed it to int and 
everything compiled fine, I didn't think much of it.  However looking 
again, the function is what ACCEPTS connections in the first place I 
believe.  

int icq_SocketAccept(int listens, struct sockaddr *addr, socklen_t 
*addrlen) {
  int s = accept(listens, addr, addrlen);

  icq_SocketAlloc(s);

  return s;
}

In the headers I have, accept is defined as:

SOCKET PASCAL accept(SOCKET,struct sockaddr *,int *);

Hence why I changed socklen_t for int.  However I guess this wasn't a 
good thing to do.  

However I had the same tcp problem with icqlib1.0 and don't remember 
any socklen_t changing in there.  Or in fact changing anything.  I 
guess it's a winsock problem if it's working ok for you on a unix based 
system. 

David

Re: [icqlib-users] TCP problems.

[Bill S. <we...@ri...>]

On Wed, 31 Jan 2001, David B wrote:

> Hi, I'm having a problem with the tcp part of icqlib.
> 
> I can do everything on the UDP side of things, receive msg's, send
> messages (through the server) etc.  However if I enable tcp, messages
> sent to me just time out and never appear.  However if I message the
> person (ICQ_SEND_DIRECT) then I can get msg's from the person.
> 
> I have the socket handler stuff set up and it seems to be working ok.  
> I also have the same problem if I call icq_Main on a 100ms timer.
> 
> Have I not set up something vital?  

It sounds like you're doing everything correctly.  The socket handler
callback is only there as an option - icqlib should work the same as it
used to if you're using the timer method to call icq_Main, it's just going
through my new socket manager code.

I've been getting a number of different reports for this lately, but on my
local system here, everything appears to work fine in both directions.  
If you could, please do the following:

1. Ensure you've updated to the most recent CVS version - I put a few
fixes in for the socket manager code after my initial checkin.

2. Use the icq_Log callback to capture a log.  Log everything that comes
across the callback, try to send a TCP message to your program.  Send me
the resulting output.

3. Give me a quick description of your system, running uname -a at a shell
prompt is fine.

Thanks,
Bill

[icqlib-users] TCP problems.

[David B <de...@so...>]

Hi, I'm having a problem with the tcp part of icqlib.

I can do everything on the UDP side of things, receive 
msg's, send 
messages (through the server) etc.  However if I enable tcp, 
messages 
sent to me just time out and never appear.  However if I 
message the 
person (ICQ_SEND_DIRECT) then I can get msg's from the 
person. 

I have the socket handler stuff set up and it seems to be 
working ok.  
I also have the same problem if I call icq_Main on a 100ms 
timer. 

Have I not set up something vital?  

David

Re: [icqlib-users] icqlib

[Bill S. <we...@ri...>]

On Sun, 28 Jan 2001, Roland Rabien wrote:

> I have been using icqlib and having no problems, today I upgraded to the
> latest version of the code (i hadn't upgraded for maybe 4 months) and I
> changed my code to fit the changes to the API, and then when I try and
> connect it crashes on this line:
> 
>   memset(link->d->icq_UDPServMess, FALSE, sizeof(link->d->icq_UDPServMess));
> 
> in icq_Login()
> 
> link->d is null
> 
> what am I doing wrong? (i am using win32)

Are you still allocating your ICQLINK struct manually?  The newest API
requires you to use the new icq_ICQLINKNew function to do this - which
will set up the icqlink->d pointer for you.  See the CHANGES_SINCE_1.0
file for more details.

Bill

[icqlib-users] icqlib

[Roland R. <rr...@gu...>]

I have been using icqlib and having no problems, today I upgraded to the
latest version of the code (i hadn't upgraded for maybe 4 months) and I
changed my code to fit the changes to the API, and then when I try and
connect it crashes on this line:

  memset(link->d->icq_UDPServMess, FALSE, sizeof(link->d->icq_UDPServMess));

in icq_Login()

link->d is null

what am I doing wrong? (i am using win32)

Thank You