Hello Anthony!
I'm trying to setup a very simple 2 system "network"... Both ends are using ngircd (I'm using 0.7.5, the other end is using 0.7.0) I keep getting a rather endless loop of the following in my /var/log/messages:
Dec 23 19:54:21 melvin ngircd[12231]: Accepted connection 1 from 12.241.198.87:49787 on socket 4. Dec 23 19:54:21 melvin ngircd[12231]: Peer announces itself as ngircd-0.7.0 using protocol 2.10/IRC+ (flags: "CZo").
Okay, the connecting server announces that it supports link compression (the "Z" flag) ...
Dec 23 19:54:21 melvin ngircd[12231]: Server "ls.xj220c.net" registered (connection 1, 1 hop - direct link). Dec 23 19:54:21 melvin ngircd[12231]: Enabled link compression (zlib) on connection 1.
... and therefore this server enables it.
Dec 23 19:54:21 melvin ngircd[12231]: Server "ls.xj220c.net" unregistered (connection 1): Decompression error! Dec 23 19:54:21 melvin ngircd[12231]: Connection 1 with 12-241-198-87.client.attbi.com:49787 closed (in: 0.1k/0.1k/120%, out: 0.3k/0.3k/81%).
The "Decompression error" message is generated by the Conn_Close() call at src/ngircd/conn-zip.c, line 165. But the more informative syslog message with severity "ALERT" which contains information about the internal state of the compression code is missing here. Is your syslog configured to direct such "alert messages" to another locatin?
any ideas why it's having an error so quickly?
The call to inflate() at src/ngircd/conn-zip.c didn't return with Z_OK but an other result code (which should be logged in the missing line of your log).
Which systems are involved? And which versions of the zlib are used?
Regards Alex