FAST 3.2.2c (2009-02-03) for TMN-F/TMU-F

[svens]

I'm not new to computers, or FAST.

I didn't want to imply that, it's not what I meant, sorry.

does the version get reported to the database when FAST starts up,

This part let me think that you didn't restart your FAST .

Just had a look at the code (because there's good documentation sigh). The server information (tool+ver) is sent when FAST connects to the database, which happens in the init function of the database plugin (=> FAST startup).

[Mort]

Maybe a bug somewhere then. In-game all server report as 'w' in the server info and when you connect, and the only difference between the servers reporting the correct version and the others is that the 'w' servers have had people playing and getting dedi records since the upgrade.

[Slig]

That's very strange yes, and i never got it with my own servers. btw it's harmless. Probably something on Dedimania server scripts, i will look at it later...

[Mort]

Thanks slig, no rush. As you say it is harmless.

Hope you got the copper donation I sent you. Might not be all that much but we're only a small group of disorganised people

[Slig]

Two out of 7 of my servers are now showing as 3.2.2w. I'm just wondering, does the version get reported to the database when FAST starts up, or do only get the version when a new dedi record is sent to the database?

In some unusual conditions some server values were not updated as wanted on Dedimania database. It should be fixed now.

[Mort]

Yep, all good. I noticed they all updated while I was doing some work on the server last night.

[JumperJack]

small question, but for some reason the finish timeout on my server is 40 seconds long everytime. i check the matchsettings, and it was still on "1". any idea why this suddenly changed?!

[Slig]

small question, but for some reason the finish timeout on my server is 40 seconds long everytime. i check the matchsettings, and it was still on "1". any idea why this suddenly changed?!

Did you checked the finishtimeout value on the server (using /info for example) ?
Also, in what mode are you playing ? in Rounds newrules mode i think that the finish timeout is based on the author time or something like that

[Mort]

=1 is the adaptive timeout, so it makes the timeout a percentage calculation of the time taken to finish the track. I can't remember the exact value though.

Did you want it to be a specific amount? If so set it to something other than 1, but keep in mind it's usually nice to give the other racers a chance to finish as well.

I use 1 myself as it means for short tracks you get a smaller timeout, but long tracks get a longer timeout for the people that aren't as quick on it.

Another thought. Are you using servermania or GPP or another plugin? They might also be setting the timeout value and overriding what you have set in the server config.

[JumperJack]

no, i'm manually altering all settings, and it's fixed again, the day after it was. i remember that a teammate of mine had the exact same problem at the same time. dunno what it is, but it disappeared automatically.

[Mort]

Hi slig,

I've got that issue with the AutoQueue and the newer FAST versions causing the message about idlekick to keep coming up even when the person is no longer on the server
viewtopic.php?p=175267#p175267

Any thoughts on what to look for, or what logs/debugging I can do to help work it out?

[Slig]

Hi slig,

I've got that issue with the AutoQueue and the newer FAST versions causing the message about idlekick to keep coming up even when the person is no longer on the server
viewtopic.php?p=175267#p175267

One of my servers with Autoqueue is constantly putting up chat messages every 30 secs or so saying "Kicked idle player at 6:00 minutes"
In the FAST console I see the messages showing two players it is trying to idle kick, but they aren't even on the server. Has there been a change to how FAST handles idle players that has the AQ confused?

Any thoughts on what to look for, or what logs/debugging I can do to help work it out?

Humm... because it is multi script, autoqueue is not a classic Fast plugin : it does not use Fast arrays to get players states etc., but get/build/maintain is own infos, so nothing in Fast should have modified the autoqueue idle kick.

Can you give more infos about your console messages ? also, when it happens can you try a '/debug autoqueue' to see what is in the [playerList:private] => Array ? players that are not on server should just not be in the array.
Autoqueue rely fully on PlayerDisconnect to be received, so perhaps that in some unusual cases the event is missed or dropped and so the player not removed from array.

You can try to add that just after $this->addCall('Kick', array($login)); in doIdleKick() function of autoqueue :

Code: Select all

			if ($this->removeFromQueue($login)){
				$idx = $this->getQueuePos($login);
				if($idx >= 0)
					$this->notifyQueue('', 0, $idx);
			}

or perhaps just that :

Code: Select all

			$this->removeFromQueue($login);

And/or try using Fast 3.2.2y , your issue should be fixed (i hope)

[Mort]

The only messsages I have are the same as what I posted (see link in previous message).

When it happened again I'll try the debug thing to see what I get, then I'll try the new FAST.

It also occured to me that there was a TM server update around the same time I updated FAST, so it might be related to the Server update, not the FAST update. I'll try reverting the versions of both and see if I can replicate the issue more readily.

[shaft78]

I would first like to say thank you guys for all the work you've done on Fast. I've been using Fast for quite some time now on my server and i like it very much.

I have just updated to 3.2.2y and so far no problems, although i didn't install the autoque this time. I was having some issues with it while using 3.2.2c, mostly people would get stuck in que and the onscreen "click here to join" didn't work at time's.

I"ll post if i have any problems with the new version, thanks again and keep up the great work

[Mort]

I actually disabled the autoqueue yesterday and just now realised I hadn't tried updating to the "y" version. Just going to update now and see if it sorts the problems out.