Talk:Server queries

From Valve Developer Community
Revision as of 17:07, 26 November 2008 by BarkerJr (talk | contribs) (A2S_SERVERQUERY_GETCHALLENGE not working since last HLDS update)

Jump to: navigation, search

SteamAppID

@Megaoerti: The change with the bot count and the unknown value can't be correct :/

Yes you're right, but i worry about the SteamAppID set as a 2 byte value 'F000'. This isn't zero?!

F0 is in decimal 240, which stands for CS: Source (see Steam Application IDs) --King2500 14:53, 15 Aug 2005 (PDT)
ok ... i forgot the sweet little endian of these two bytes *g* --MEGAOerti 13:17, 16 Aug 2005 (PDT)

Protocol

Source seems to handle multiple packet replies differently. It also sends the 0xFFFFFFFE header followed by the request id. But then the next byte contains the total number of packets, the one after the current packet number. Also Source does not seem to start all "actual" replies with 0xFFFFFFFF (only contained the first packet).

Is this completely right or did I miss something? If it is it will have to be updated in the main page.

--Philip 08:41, 4 Sep 2006 (PDT)

SteamAppID Again

Either I'm being blind and stupid, or do Goldsource a2s_info replies not include an AppID... I think I should point out that I may be both blind and stupid, but I'm just not seeing it

LINE Hex
0000 FF FF FF FF 6D 31 39 34 ....m194
0008 2E 31 34 30 2E 32 34 32 .140.242
0010 2E 32 3A 32 37 30 32 35 .2:27025
0018 00 32 5E 48 41 42 20 43 .2^HAB C
0020 6C 61 6E 20 53 65 72 76 lan Serv
0028 65 72 3A 3A 77 77 77 2E er::www.
0030 32 68 61 62 2D 70 72 6F 2hab-pro
0038 67 61 6D 65 72 73 2E 63 gamers.c
0040 6F 2E 6E 72 3A 3A 00 63 o.nr::.c
0048 73 5F 6F 66 66 69 63 65 s_office
0050 00 63 73 74 72 69 6B 65 .cstrike
0058 00 43 6F 75 6E 74 65 72 .Counter
0060 2D 53 74 72 69 6B 65 00 -Strike.
0068 01 0B 2F 64 6C 00 01 77 ../dl..w
0070 77 77 2E 63 6F 75 6E 74 ww.count
0078 65 72 2D 73 74 72 69 6B er-strik
0080 65 2E 6E 65 74 00 00 00 e.net...
0088 01 00 00 00 00 9E F7 0A ........
0090 00 01 01 00 ....

--Luthias 09:05, 13 Nov 2006 (PST)

You're right. GoldSrc server replies doesn't include an AppID. --King2500 17:32, 11 Jan 2007 (PST)

Counter-strike server reply format

I couldn't find anything about CS1.6 servers, AFAIK this is their reply format for A2S_INFO. There is more after this but I dont know what it is.

Data Type Comment
Type byte Should be equal to 'm'
IP string The server's IP address and port
Server Name string The server's name, eg: "Recoil NZ CS Server #1"
Map string The current map being played, eg: "de_dust"
Game Directory string The name of the folder containing the game files, ie: "cstrike"
Game Description string A friendly string name for the game type, ie: "Counter-Strike"
Number of players byte The number of players currently on the server
Maximum players byte Maximum allowed players for the server
 ??? byte A '/'
Dedicated byte 'l' for listen, 'd' for dedicated
OS byte Host operating system. 'l' for Linux, 'w' for Windows
Password byte If set to 0x01, a password is required to join this server
Secure byte If set to 0x01, this server is VAC secured
 ??? string "www.counter-strike.net"

More information to add to this article

The French version of this article now has more information than this one (the english version). I have asked the person that is currently updating the French article if it were possible to add the same information to the English version. If anyone else has the time to do it as well, please step forth! Perhaps we need a new template to "tag" situations such as these... a template like {{translationupdate}}, but for any language (specified when the template is used), rather than just for the English version. --Etset 14:23, 31 Jan 2008 (PST)

Bug on Rag Doll Kung Fu server reply

Sometime (actually once in two) a RDKF server will reply like this :

FF FF FF FF 49 FC 55 00 53 00 45 00 52 00 20 00    ....I.U.S.E.R. .
32 00 27 00 73 00 20 00 64 00 6F 00 6A 00 6F 00    2.'.s. .d.o.j.o.
00 00 53 6F 63 63 65 72 00 52 44 4B 46 53 6F 63    ..Soccer.RDKFSoc
63 65 72 00 52 61 67 44 6F 6C 6C 4B 75 6E 67 46    cer.RagDollKungF
75 3A 20 53 6F 63 63 65 72 00 EA 03 01 04 00 00    u: Soccer.......
77 00 00 32 2E 33 2E 30 2E 30 00                   w..2.3.0.0.

Everything is fine except the server name (in blue) : a NUL byte is inserted between each character. There is the same effect with the player names on the A2S_PLAYER query. What about? --Cortexd 17:55, 17 Feb 2008 (PST)

I think these strings are in UTF-16 instead UTF-8 --Cortexd 19:49, 2 Mar 2008 (PST)

GoldSource RCON Protocol

Hi all,

This is probably an old topic - but is there any definitions for "GoldSource" aka "Classic" rcon packet requests/responses. Is this the correct place to start one ?

I am slowly working my way through the protocol so I may be able to contribute in some areas as well.

cheers will Beattie Wlbeattie 19:04, 6 Mar 2008 (PST)

Player connection time

Does anyone know what a player connection time of -1 (0x000080bf) stands for? Maybe someting like "is connecting"? --Koraktor 23:55, 7 Aug 2008 (PDT)

Bots I think. --TomEdwards 09:26, 8 Aug 2008 (PDT)

A2S_SERVERQUERY_GETCHALLENGE not working since last HLDS update

GoldSrc servers do not respond to A2S_SERVERQUERY_GETCHALLENGE requests since the last HLDS update (GoldSrc protocol version 48). Does anyone know what has changed? Or is this a bug? --Koraktor 07:42, 2 Nov 2008 (PST)

Indeed, kept me busy for quite some time too, but got the solution with the help of some others:

I'm using the function below. If you look closely you discover that the 5th byte has changed to 55 instead of 57 and it's now followed by 4 times \xFF .... dunno how, why and so on, but it works!

function getQuery($queryType) {
	switch($queryType)
	{
		case "A2S_SERVERQUERY_GETCHALLENGE":
		return "\xFF\xFF\xFF\xFF\x55\xFF\xFF\xFF\xFF";
		break;

		case "A2S_INFO":
		return "\xFF\xFF\xFF\xFFTSource Engine Query\x00";
		break;

		case "A2S_PLAYER":
		return sprintf("\xFF\xFF\xFF\xFF\x55%s", $this->getChallenge());
		break;
		}
	}

Hope this works for you too  :) —The preceding unsigned comment was added by Ouwe tosti (talkcontribs).

"Does anyone know what has changed? Or is this a bug?"

Current protocol 48 bugs:

  • A2S_SERVERQUERY_GETCHALLENGE does not work at all, use A2S_PLAYER or A2S_RULES instead with ffffffffh (-1) as challenge.
  • The map name from A2S_INFO is only updated when players are present.
  • Hostname and bot flag from A2S_INFO are only updated when the map changes.
  • Server rules from A2S_RULES, like amx_nextmap or mp_timeleft, are only updated when the map changes.
  • mp_timelimit is off by a few minutes (only the value returned by A2S_RULES, rcon is ok though).
  • bot count is now a bot flag, 1 means there were only bots during map change.

I'm not exactly sure which part of the query info gets updated when so take these bug descriptions with a grain of salt. Example, I join an empty server and the map name gets updated, but mp_timeleft is still wrong. I join a server with bots that also has a plugin that updates the server name with the time left, but only the map name gets updated via the query, not the hostname. (You can check the actual full-length hostname by typing status in console.)

--pizzahut 23:30, 2 Nov 2008 (PST)

Looks like all GoldSource servers are responding with the Source protocol format. You will have to use appids to tell source/goldsource servers apart (IE CS/CSS). --The MAZZTer 09:47, 9 Nov 2008 (PST)

A2S_SERVERQUERY_GETCHALLENGE seems to be ignored, because it is changed. It is now 0x55 instead of 0x57 (and Steam seems to send an extra Int32 0 after it, not sure what that is). --The MAZZTer 13:52, 10 Nov 2008 (PST)

The A2S_SERVERQUERY_GETCHALLENGE hasn't "changed", it stopped working with the update. Steam apparently never used it. The header 0x55 belongs to a A2S_PLAYER request. The additional (long) 0 is an invalid challenge, therefore the server responds with a S2C_CHALLENGE containing a valid challenge. --Koraktor 10:06, 13 Nov 2008 (PST)

I put a warning on the main page about this, but it'd be really nice to know if this is a permanent change. Also, it'd be nice to know if this query will stop working in Source servers in the future. --BarkerJr 09:07, 26 Nov 2008 (PST)