JvIasterMind

Fair enough! I will keep my eyes open to see if any of those are related, but it happens with multiple players and one of them is a server regular (so I'm pretty sure he's not downloading maps). The lua callback would be a great addition! I would also love the notification to include info on if it is a known cheat binary, because that will get the cheater off the server much faster than if the checksum needs to be read from the server logs. Still, even with this info, I still think that the lua callback should be called. Maybe add a boolean in the callback if it is a known cheat version so it can optionally be dealt with in a different way from the unknown ones.

I have a couple feature requests that involve the checksums for unauthorized binaries (sort of on topic). -Could you guys add the checksum to the warning message that displays in admin chat for the unauthorized binary? This will allow admins to instantly ban a player if they notice a checksum that they know is malicious. Today, I had to download our server log and look through it to get the checksum and I instantly recognized it as the wallhack that I listed earlier in this thread. -Currently, the warning message shows up twice... once during warmup and once during the start of the map. Could you guys make it so it only shows up once for each map instead? It starts to get a little annoying when there is a player that isn't raising any suspicion on the server. Also, I have a bug that goes with this. Sometimes the message is not displayed and then displays the next time the message appears (so it pops up multiple times). I have even had it skip a map completely and then show four times on the next map after warmup finished. -Another feature request, (especially if you choose to not implement a config file like I suggested in my last post

If you playdead a moment before hitting the ground, you can avoid fall damage. A few players on our server take advantage of this glitch and do it every time they jump off ledges. I believe that it behaved like this even in ETPub, so it is not really anything new. I still think it would nice if it could be fixed.

I like this idea. We haven't been using the incognito flag because we normally want players to be able to tell if an admin is on the server in case they need to report issues. If there was a way to toggle it on when needed, I agree that it would be very useful for specing a player for hacks. You can actually accomplish what you describe most of the way by creating a custom level with incognito and setting your level to that one when you need the flag set, but I am not sure you can hide the !setlevel message from showing. The only problem with the renaming and setting incognito is that if there aren't very many players, someone could easily notice the name change and know who the player is anyways. One possiblility to overcome this would be to disconnect and then reconnect with incognito set (probably needs to be able to be set on the command line so one can make a custom shortcut or something).

I like the idea of a checksum config file! However, I think it would be better to allow admins to input checksums that would instantly ban a player connecting with a known malicious binary, instead of disallowing all unknown binaries (since some have legitimate uses as described by TheSilencerPL). By the way, silEnT does allow you to auto kick any unrecognized binaries similar to how you described in NxAC (minus the config file as far as I know). As for a global banlist, I have similar thoughts to Beck and feel that different servers have different tolerances for what is considered enough proof to constitute the ban. I know of quite a few players that were wrongfully banned for hacking, when in reality, they are just good players. I feel it would be terrible if a legitimate player got banned on one server and got onto the global banlist and was unable to play anywhere. If the global banlist was only for binaries that are 100% known to be malicious, than I would be fine with that, but I would want to stay away from as much human error as possible.

Ok, got it! Thanks for the info!

What do you mean by this? Should we assume all unknown client binaries are safe until we see evidence proving otherwise? Or are there legitimate reasons why a players would be using an unknown client binary?

Not to derail this thread too much, but what are the chances of the other unknown client binaries being safe? We have a few players that report the error message with other checksums that googling doesn't lead to anything with (their playing has not been too suspicious). I agree with Beck that it would be awesome having some sort of database of different checksums so we can look up players to know if they are hacking or known safe (or even ones that are not determined).

Caught two different people using the following binary. Google searching that checksum leads to a wallhack. 30d13c5e1339f3c80ad89ab4525564ab

I still like the idea of splitting the bans out of the shrubbot.cfg file. Right now, it seems it is not clear enough that bans are placed in there. Currently, our shrubbot.cfg file on our server is already pretty big without the bans in it (we have quite a few levels defined and also a bunch of custom commands). Actually, now that I think of it, wouldn't it be better to split all the info out of the file so we have files like levels.cfg, commands.cfg, and bans.cfg?

I have another feature request that is related to this one. What about having an option for enemy artillery strikes that are within a covert ops' line of sight to show on the team's command map as well (but maybe red instead of green)? I feel like covert ops is the least utilized class in the game, and this would help add extra utility to the class.

I like this idea! Do airstrikes show up on the map already (without the radius)?

Maybe there is a problem with the clients getting the file. Does the client download the file automatically when you connect to the server?

I think it would be cool if death sprees were added to the spree field of the Kills Display. Here are two methods that I think would work good if this feature was to be added... Use negative numbers for deaths in the spree field and just keep it the current color. This would be my preference because I like the contrasting colors that the counter uses currently. Have a different spree color for death sprees. Maybe make sprees just be a more dull version of the kills and deaths fields to indicate which one it is.

Thank you very much gaoesa!!! I'm really looking forward to this addition!

On our server, we are just using custom sounds.

This sounds like an interesting idea! I don't know much about those .voice and .sound files. I think it would be best to keep the configuration for server admins as simple as possible, which means to use the keywords that are already defined or setting a cvar like most other settings.

Ok, all that makes a lot of sense. Thanks for answering all my questions! Also, thanks for thinking about and considering the idea! I really appreciate all the work you put into the mod and how great your support is in these forums!

Hmmm... I guess I didn't think about the network part of it! I have a few questions if you don't mind answering them... (Keep in mind, I am not that good at coding, so the following questions might be just a bunch of nonsense. ) Doesn't silEnT have to read all the client cvars when they connect to the server and also when they modify them to make sure they comply with the cvar restrictions? Could the server just keep all the cvars in memory for the connected clients (or does it already)? What about only allowing the accessing function to be during the et_ClientConnect or et_InitGame callbacks so the time it takes to get the value wouldn't be as critical and maybe there could be some kind of timeout where the value would return nil or something? Also, there is the Userinfo string with some cvars already. Could an alternative to that be implemented in much the same way with different values (or certain cvars added to that string)? Most of the values in that string seem like ones that wouldn't change during game, but I do see that it has a name field. How is that field kept up up-to-date? Maybe the same could be done with cvars with a set prefix like lua_ or similar?

That is an interesting workaround. Thanks for posting the code... very fun to see how other people accomplish similar scripts!

Anybody know the answer? I tried searching to see if you can get client cvar values through rcon and haven't had any luck. It seems that rcon is not documented very well in the ET community since many of the sites that used to be good are offline now. If it isn't currently possible, maybe it would be a good feature to add to the lua api. One thing to consider, however, is do you think that being able to read all cvars could potentially pose a little bit of a security risk for clients if they have there private password or rcon password set on there command line in their ET shortcut? If that's the case, a couple solutions could be to read all cvars starting with "lua_" or just keep any security related cvars disabled. Otherwise, it would be pretty useful in some cases for lua to be able to read all cvars. An example of this could be on a limited lives server, lua could display a message to any client (shortly after connecting to the server) that has cg_instanttapout set to 0, to change it to 1 if they don't want to see the limited lives popup menu. Additionally, this message could even be shown right after a player dies and they see the menu for the first time! The main way I want to use it currently is to have some configuration options available to the clients. For example, I am working on a script that plays custom sounds on our server. I would like to give the client the option to disable this functionality if they do not want to hear them. The way I am currently planning on doing this is to allow the clients to set a custom cvar such as playsound_disable, so they can just add the line "set playsound_disable "1"" to their config if they want them disabled on our server. This would be excellent functionality because setting cvars is mostly common knowledge in the ET community, and I would be able to avoid workarounds such as having to make players execute a custom command and recording the preference in some type of data file on the server.

I like the bot filtering idea! Another possible filter that could be interesting is one that filters out servers that just redirect to other servers. It seems there are some servers set up that are just "dummy" servers that all redirect to one server. I feel that doing something like this tricks players (especially if the dummy servers have a different ping that the target server).

Is there a way to get a client cvar value through lua (similar to cvarvalue = et.trap_Cvar_Get( cvarname ))? I didn't see a dedicated command, so perhaps it can be done through et.trap_SendConsoleCommand( when, command )? What I would like to do is allow clients to configure some settings related to lua scripts I am writing. I figure that the client setting some cvars for configuration and having the lua script read them would be a good way to do this.

You where correct. I had a bug in my code that was making it appear like this wasn't possible. I've fixed it now and it works perfectly. Thanks gaoesa!

Anybody? I would have thought that there would be a way to modify the times through the server console, but I am having a hard time finding any info on how to do this online. If it isn't currently possible, would the silEnT devs consider adding a function to be able to modify these types of variables to the lua api? I'll give a little background on what I am working on. I'm trying to use lua to decrease the limbo times of every map by a universal factor. I want it so we don't need to use customized mapscripts every time we add new maps to the server (especially, since we tend to cycle our maps quite a bit). Modifying the g_redlimbotime and g_bluelimbotime cvars during the InitGame callback works great except for maps that change the limbo times in the mapscript for objectives such as building a CP. I would like to be able to decrease those objective modified times as well to keep everything consistent.