I have this issue where my immortals (my 60 included) are losing link after about 9-10 minutes of non-action. Now at first thought, you would check update.c for the char_update function. Of course everything seems fine there, and I have even thrown in a couple of IS_IMMORTAL checks to make SURE that this isn't the problematic issue. Now. When my connection to my host disconnects me due to inactivity (in the shell), the mud seems to just up and quit. However, if I exit the shell, the mud will stay up, but the immortal I am on will STILL disco after about ten or so minutes, even though the mud continues to run. Any ideas on this?
yes the net death message, which is the generic message you get when you lose link.
Could you be more precise?
It sounds like the MUD is disconnecting you for being idle when it shouldn't, and the Linux shell is disconnecting you for being idle. If we can establish that losing link with the MUD is in fact the server, or some router in between, cutting you off, that's quite different from the MUD timing you out. Therefore knowing the message you see in your client is important.
One thing that can be easily recommended off the bat is to try using TCP keepalinves if your client supports those, in case this is indeed a router somewhere along the connection cutting you off of inactivity.
I am indeed connected via router (on a lan I have set up in my house), so that very well may be the issue as far as the imm just getting disco'd. David, the exact message is: Net death has claimed suchandsuch, and it's found in comm.c, in the close_socket function.
Is that the message that is sent to the character that's being disconnected, printed to the logs, sent to other imms as one imm is being disconnected, printed by the MUD as it disconnects a character, printed by the MUD as it realizes a character's connection is down, …… ? Basically all I'm trying to establish is if the MUD is initiating the disconnection, or if it's simply observing that something else killed the connection.
If it – the MUD – is just disconnecting idle connections, that doesn't not explain why the shell is also disconnecting idle connections. All indications so far are that it's not the MUD doing it. I think that Runter is correct: it's the "whoops, that connection just died" message, but I'd like confirmation.