gaoesa

et.trap_Argv(0) = the actual command, in this case "lua" et.trap_Argv(1) = ammo et.trap_Argv(2) = [1]

It is possible I think. I would need to confirm it from the source codes for certain answer. But I don't think there are major problems with that.

Parameter substitution should also work. [command] command = ammo exec = lua ammo [1] desc = bla bla bla levels = 0 1 2 3 4 5 The benefit of this approach is that you can use the inbuilt admin system to control who has access to the commands and the commands also work in the exact same way as inbuilt shrubbot commands. I.e. no need to handle admin levels, console writes and chat writes in scripts. The downside is that there is some amount of replication and dependencies between shrubbot.cfg and Lua scripts. That should work. Let us know how it works out.

That is a good suggestion. The major problem with the current documentation is that it is describing the differences to the ETPro's documentation. http://wolfwiki.anime.net/index.php/Lua_Mod_API

The custom commands are an existing feature. http://mygamingtalk.com/wiki/index.php/Silent_Shrubbot#shrubbot.cfg That is precisely what I was thinking. I'm not sure what you mean by runing Lua scripts through rcon, but there is no rcon command to execute scripts. However, the script that is loaded on map start can be triggered to do things.

I was thinking something like this: [command] command = example exec = mycommand param1 param2 desc = Invoking Lua from custom commands levels = 0 1 2 3 4 5 And in the Lua script somthing like function et_ConsoleCommand() -- check if it is a command to intercept -- if so, read parameters and execute the associated Lua functions end About the ClientNumberFromString, it seems to be the same thing with the ETPub. I'm wondering if it should be fixed to match the documentation or if documentation should be fixed to match the current function name. It depends about which syntax is more used and changing it would cause bigger problems.

Would it be possible to use Lua calls in shrubbot custom commands? Maybe trapping server console commands in Lua script? EDIT: It looks like the et.G_ClientNumberFromString would be actually just et.ClientNumberFromString. Does that work?

You can use the map configs system with any settings and commands. http://mygamingtalk.com/wiki/index.php/Silent_Mod_Server_Cvar#g_mapConfigs

It does look like that the crash doesn't happen within the mod. There is nothing we can do, unless he can figure it follows some specific action.

Yes. But it is actually silent_crash.log. Sorry about saying .txt first.

According to the log, the server is receiving signal 2 which means terminal interrupt. Do you have some cronjobs that are working on the server or something? That is not a crash though.

I think safe numbers are anything below 100. Maybe more. Maybe you could try 20 or 50 if it fits your purposes. It of course depends with the frequency the players on your server will change the nicks. Some players never change them. On the other hand bots do rotate the nicks so don't get scared because it looks like it is growing fast in the beginning. If you set the g_dbMaxAliases to a smaller value, but still more then 0, it will remove the older aliases in the next server close. If you set the g_dbMaxAliases to 0, nothing will happen to the database, but it is not used. Do remember that this requires server restart.. The example I mentioned doesn't need SQL at all. I remember there was some sort of clan tag protection script and I think quite possibly this could be modified to store single informations about a player. E.g. designated name. If you want to use SQL through Lua there are possibilities for it. But you need to use the qagame binary that doesn't have Lua statically linked into it.

The reported issue was client crash. Server crash is unrelated to client crashes. I know it reads empty in the client crash information dialog, but could he still check from the silent folder if there is data in the silent_crash.txt file. Bacteria, could you post the server crash information? It can be found from the console log and the server log both.

The alias inserts happen in constant time and the commands !aliases and !userinfo will not have significant performance losses from bigger aliases lists. So there shouldn't be problems experimenting higher g_dbMaxAliases values. The only times when the performance suffers from bigger aliases db is when the db is cached into memory during the map initialization and when using !aliasessearch command. I do think that hiding the true identity can be done by changing GUIDs in general. Also, I do think players should have the option to change it also. At least over time. Using the example of small g_dbMaxAliases value and dropping the least used aliases, it would be possible to hide short lived offensive nicks and fake being a nice regular. While in the case of dropping the nicks by time, even if the most used nick would get lost it would be possible to determine from other stored data that the player is supposed to be a regular. Another poblem is that if the player has 2 nicks with a lot of playtime, it will at some point take considerable effort to have a new nick overcome either one of those old ones. Effectively he would have a possiblity to change his nick freely and all those changed nicks would be lost. We can consider adding option to remove the least used nicks. PS. I think if you want to set some sort of identity info by hand to a player like in the !setlevel example above, I think a Lua script might do that easily. Associating one name, or note, with one guid.

It drops the oldest aliases. You can limit the name changes per map with http://mygamingtalk.com/wiki/index.php/Silent_Mod_Server_Cvar#g_maxNameChanges

The problem with this is that the weapon selector is made with static menus. While in silEnT mod the weapon selections are very configurable. If you set g_classWeapons to 0, this menu will work correctly, but depending how the setting is, the menu will have incorrect numbers.

It's a server configuration option http://mygamingtalk.com/wiki/index.php/Silent_Mod_Server_Cvar#g_inactivityOptions Flag 2 and followers are not dropped.

To be exact, there was 4 grey stars after each skills white stars. Making the maximum 8 if the skill was full.

Found and fixed in 0.6.1. This was graphical bug related to the skill_(x) changes in 0.6.0. Thanks for reporting. Please report more if you notice any.

There seems to be an alignment bug with the stars. The 4 last stars from the 8 are probably the other skill on the right side. Do you remember this bug being in there in previous versions as well? This will be addressed in the 0.6.1.

They haven't been changed in any way. What is the exact complaint? Remember that you can test the boxes locally with g_debugHitboxes cvar. There is also g_debugBullets cvar.

These are the database files in the g_dbDirectory. The mod will do it automatically though. Old db files can be allways found with _v(xx) attached in their names.

Thing to note is that in this version the database will convert to a new version. If you wish to play safe, you can safe copy userdb.db and userxdb.db files. If you want to use the new aliases database, you will need to set the g_dbMaxAliases to positive integer before you start the server. This setting cannot be changed on a running server. If your players have mutes before the database is updated, those mutes will be cleared because of the new mute format that is both tied to real time and includes information about who muted and what reason was put for it. Another feature to note is the skill_(x) cvars. You can now enable/disable any skills regardless of the order. With 0, enable and -1 disable.

The doanload is temporarily unavailable because a wrong version of the client pk3 file. It will be made available again once it is fixed.

http://mygamingtalk.com/wiki/index.php/Silent_Mod_Server_Cvar#g_excludedMaps In this case it would be, if the map names are right, set g_excludedMaps ":oasis:goldrush:"