multiplayer concerns

A forum for online playing, administration, bugs and feature requests
Guest

Post by Guest »

ace123 wrote:
YetAnotherGuest wrote:Here are some anti-cheat strategies coming to my mind, while reading this thread.

Let the server only route data to the clients and let the clients check for cheats.
Every client needs to know movement- and positionvectors for all visible objects to draw the correct environment. Also object-types need to be known to draw the correct models.
With this information the client also knows the maximum parameters of other objects (e.g. the maximum armament -> maximum weapon power, also maximum movement parameters).
If the clients detects any irregular data coming from other clients, it can inform the server.
The server then checks the data coming from the supposed cheater and if the client is found to be cheating it can be kicked.
With this method you can use a space-saving protocol for network transfer and also limit server cpu usage.

Nevertheless this method will not stop cheating, but will limit cheats to the maximum capabilities of ones ship (which can also be achieved regularly).
First, this would force the client to display all units and know about every system. Otherwise a player could go into an empty system, cheat and come back into a populated area.

Also, a cheating client could still make every stat exactly the upper limit of everything without anyone noticing, which would still be bad.

It's quite possible that a cheating client would find a flaw in the algorithm to check if a ship is cheating or not and it might get the best ship while inside the limits of not cheating.

It's also possible that an innocent client performs an upgrade and either due to some rounding errors or a bug in upgrading, the client goes outside of the upper or lower limits for a ship and is considered cheating and thus kicked out. This would likely result in a very unhapy and unsatisfied player who would likely not play Vega Strike again.

Not to say that the server dealing with all th data method cannot alow for cheating, but really how much can you do if you can only send keyboard and mouse events to the server. You may be able to do some actions faster in a cheating client but likely it won't make a difference.
Also, some people could still write an AI for a human player which acts better than the human and always hits and sends the keyboard events to the server to make it look like a real player.
Standalone AI bots wouldn't be a bad thing, it would be cool to be in compitition of who can make the best AI and then send them to war.
captnemo
Merchant
Merchant
Posts: 51
Joined: Mon Aug 25, 2003 9:34 pm

Post by captnemo »

Instead of doing all calculations on server, you could just do spot checks, verifying so many calculations per second, adaptable to server speed. This wouldn't stop aimbot type cheating, but aimbots wouldn't be anymore effective than a good autotracking system.
Alask

Most solutions...

Post by Alask »

Ok,
So far these arr the things that I have seen highlighted

Re comiling with a different client and
AIMBot style hack.

The different client needs to have two solutions
You could send a seed and ask the client to compute a hash using the files and the seed (Stopping the client from just caching the correct response) but then they can just make a copy of the folder and use that as the file source.
Another part is to hash the memory values with a seed to stop hacks like that.
So we can stop most people from using an unpure client. But they can still have a pure and unpure and do the checks on the pure one while using the unpure one. What if we scanned the program in memory, but then someone could just make a memory dump of the program.

So there isn't really anyway to stop someone using an unpure client.

It is impossible to stop an aimbot if you can't verify that the client is pure.

I havent given it much thought, but the only solution seems to be a Punkbuster style system, and it hasd to be closed source.

Just my 2c
Silverain
Expert Mercenary
Expert Mercenary
Posts: 984
Joined: Thu Aug 07, 2003 5:35 am
Location: Brisbane, Land of Oz
Contact:

Post by Silverain »

<curious>

Please correct if I'm wrong.

My understanding of the GPL is that there must be access to the sourcecode - it says nothing about being able to modify the compiled program itself!

As such, I don't believe there is anything stopping us from developing a locked version of the program (how, I don't know, I'm not a developer) that is used for multiplay.

We still satisfy GPL as the source code for the program remains available, but I as a hacker can't go and change the MP version program. If I do, presumably I change something (checksum??) that can be marked by the server.

IF this was possible, may require some rearrangement of data - I don't know - but its a possible thought and option.
THOUGHT CRIME! [points finger] THOUGHT CRIME!
Alaskan
Hunter
Hunter
Posts: 79
Joined: Fri Sep 10, 2004 12:16 pm
Location: Australia
Contact:

Post by Alaskan »

It seems the only way to stop cheating and not have a huge server is the semi p2p method.

You have a central server (or a few with a single gateway) and this server stores the current universe. This server also keeps track of the players, but if the size of this becomes too large, only a hash of the players info is needed(MD5, SHA) and the player can store their own info.

So we have the universe and the players.

Now we need to coordinate everything, so we have the central server keep a list of everyone connected, then when someone connects the server directs them to another player, who has a high bandwidth connection, and the user connects to them.

Each system needs to have a tight network of nodes and these nodes work out if the clients that are connected to them are behaving legaly, otherwise they don't allow them to do what they are trying to do. And while this is happening, all the legal info is being sent to the other nodes.

Now we have a scalable p2p network that uses a central server to coordinate everything. One problem is that most people don't have unlimited megage, and I think the nodes will use alot of that up. So maybe just a cluster of computers that each deals with one section of the universe.

It just seems that the only way to stop cheating is to have legality dealt with on the server. And then this server witll have a huge traffic bill.

I dont know, anyone else have a better idea?
2 133t for joo!!!
scheherazade
Developer
Developer
Posts: 427
Joined: Thu Jan 09, 2003 6:03 am

Post by scheherazade »

my 2 c


almost everything client side (like warbirds, so you can have a LOT of players)

client only tells the server <position> <direction> <speed> <firing t/f> <who is damaged>

yes, damage is inflicted client side... so that there is no computation to be done on the server, that affects the clients gameplay.

----

AND , ez to cheat no?... 'no'. if you do one thing.

have the server run its own version of events. it then samples if what is happening is within the realm of possibility.

i.e.
did you say you're going faster than you can?

did you damage a unit you weren't facing at some point during a firing condition? (use an "un-lagged quake"-ish method, go back in time and compare if it was true <used in quake to compare to the highest latency, when allowing client side damage>)

are you hitting 2 people who are more than some high degree apart simultaneously with guns?

are you selling more cargo than you collected?

etc.

the server corrects position inconsistencies by sending back adjusted positions, -- this is for when the position is off by some super tiny amount due to possibly differences in floating pooint or random calculations. this can be done intermittently.

server _DROPS_ people who have severe inconsistencies.

cheater bye-bye.

----


you can have smoothe gameplay, EZ aiming even when ping is crap (300-400), and you can't cheat.

-scheherazade
jackS
Minister of Information
Minister of Information
Posts: 1895
Joined: Fri Jan 31, 2003 9:40 pm
Location: The land of tenure (and diaper changes)

Post by jackS »

scheherazade wrote:and you can't cheat.
or, to be pedantic, you can't expect to cheat for very long or very many times without the odds of being caught becoming very high.

There has been precious little doubt in my mind for some time, ever since a conversation with a MMORPG enthusiast aquaintence and fellow graduate student (as an aside, as to his MMORPG credentials, one of his former names was "Belan the noble looter" [UO fans may or may not remember, but the EBAY price was amusing] and I'm quite certain he still plays more of Planetside and other MMORPGs than his PhD advisor might entirely prefer :-) - or not - he has an interesting advisor) that a sampling based approach is the correct one, if not the only plausible one, for mitigating cheating in an open-source WAN game. One can actually extend the concept a bit further too though - not all verification need be done entirely by the trusted servers. Assuming some reasonable number of active clients, and hence a corresponding need to mitigate server load, verification tasks can be dispatched to sets of clients to be run as background tasks (note that, as the clients are not trusted, these sets should contain several randomly selected clients, such that collusion is less likely and some number of failures, however byzantine they may be, can be sustained - and that the eventual decision must still be rendered by a trusted server). One might also argue that merely dropping the offending clients is insufficient - given an account based system, one imagines that the offending account should be flagged, and the events logged, both for the regrettable event that a false positive has been registered and the user complaint must be examined, and for the unfortunate necessity of tracking repeat offenders, correlating their multiple accounts and taking necessary administrative action to deny them the ability to make gameplay worse for others.
Alaskan
Hunter
Hunter
Posts: 79
Joined: Fri Sep 10, 2004 12:16 pm
Location: Australia
Contact:

Post by Alaskan »

And a system like this could reduce the need for a huge server.

http://www.scs.cs.nyu.edu/coral/

P.S. That explains why UO was so hard on a 56k, you would run so far,then teleport back to where the PK had jumped infront of you, and then all of a sudden be dead.
2 133t for joo!!!
Fink
Just a tourist with a frag'd nav console
Just a tourist with a frag'd nav console
Posts: 3
Joined: Mon Dec 27, 2004 9:35 pm

Post by Fink »

Some thoughts from a seasoned online gamer :

1 ) Fisrtly, there is no such thing as unhackable code. If a man made it he can break it (sorry not being sexist, I appreciate many tallented coders are female... is just a figure of speach here!). So we aren't talking about preventing source modification, (which in itself prevents modding etc from being so effective), we are talking about flagging source modification for online games that require set values in code.

2 ) Rather than reinvent the wheel, why not look into existing projects such as punkbuster etc? By working with developers on these projects you get the experience, knowledge and workforce (code protection is a full-time job in itself, where are all the extra coders gonna come from else?)

3 ) Why not create 2 'versions' of VS... one is the full GPL read it and weep sourccode version, and one a pre-compiled 'official' version that is the only version allowed to connect to the VS cluster... as opposed to P2P / LAN / Solo gaming. When playing say Quake 3 in a tournament there is a required official release that must be used etc... this is level 1 protection and the 'punkbuster like system' will be able to spot exceptions by hashing etc.

4 ) Have the official version of VS require registration of the user... each binary has a liicence to register it (same way a commercial game does) but of course is free licence not charged...

5 ) The licnce/registration code uniquely ID's a player allowing the servers to ban this user if required.

6 ) Allow the server to take screenshots on clients, maybe even record a demo from them.. this is becoming more common in online games as it shows the server admins proof of video hacks etc. Another important 'admin only' option is following.. the ability to latch on to a player and see how they perform... where they go what they do and how quickly they killed the entire server with 2 shots type thing.

7 ) Game admins... create a 'police force' of trusted players who have 'extra server commands' such as kick, ban, suspend, remote demo / screen shot etc. The people will very quickly begin to spot cheaters... believe me a regular online player begins to have her / his suspicions soon enough. Equip her / him with the tools to police and you will find that a major deterrent. When I have run servers in the past (everything from HL and its mods to Q1 / Q2 / Q3 ... BF1943 etc. etc. I have seen the awesome effect of good admins on a server... nothing more soul destroying to a cheat than being publicly kicked all the time... shows them for the fools they are and they seldom come back if they believe such zero tolerance admins are about.

8 ) Rampant cheating online (and it takes no time at all for this to happen) will sure as anything kill public servers if its not constrained, of that there is no doubt.

9 ) Speak with some of the commercial game writing outfits to obtain advice... I'm sure they won't deny _all knoweldge_ to you ... cheating is a global phenomena. You may even find some of them have coders playing VegaStirke right now...

10 ) before you flame here, remember we are talking about a sustainable, low-cheat tolerence system for users prepared to adhere to it. There is no one solution and those that don't wish to be so constrained need not use this system... after all you can't please everyone.

Finally, I am willing to volunteer my time / resources into a multi-player cheat solution so if you are looking for bums-on-seats count me in. :)
Guest

Post by Guest »

3 ) Why not create 2 'versions' of VS... one is the full GPL read it and weep sourccode version, and one a pre-compiled 'official' version that is the only version allowed to connect to the VS cluster... as opposed to P2P / LAN / Solo gaming. When playing say Quake 3 in a tournament there is a required official release that must be used etc... this is level 1 protection and the 'punkbuster like system' will be able to spot exceptions by hashing etc.

4 ) Have the official version of VS require registration of the user... each binary has a liicence to register it (same way a commercial game does) but of course is free licence not charged...
you should read http://vegastrike.sourceforge.net/mission.html
and first of all http://www.gnu.org/copyleft/gpl.html :!: :!: :!:

your proposals are really evil! :evil:
following them would result in a GPL violation AND make the vegastrike mission fail!

and about taking screenshots and record demos: that's evil too! vegastrike would turn into a trojan horse! no, thanks!
D(not logged in)

Post by D(not logged in) »

Ok, first off I like the idea of having two different multiplayer versions of vegastrike.. Where the first one is more of an open ended tpace trading where you interact with players(like everquest in space). And the second one where you have pure fact pacted shooter game with the following game modes(Deathmatch, Capture the flag, Basewars, etc).

Second, I don't like the idea of having some kind of a cheat protection where it would be a total nightmare for the 56k users when they want to go online playing. It also kicks people out if they have too much accuracy for having god like skills.. If anyone wants to see how a cheat can be prevented, they can check the BZFlag(open source tank shooting game) game site. Actually, for a quake in space type game, there needs to be a join as a spectator mode especially when spectating another player in first person mode to see if hes cheating or not. Actually good admins will take care the cheaters.


Last, I don't want to see vegastrike to go commercial in anyway. It's a good thing that it won't in the future.
hellcatv
Developer
Developer
Posts: 3980
Joined: Fri Jan 03, 2003 4:53 am
Location: Stanford, CA
Contact:

Post by hellcatv »

we'll make sure VS is always free...
though if people end up selling it, rest assured that there will always be a way to get a free version that's the same (or better...i.e. more up to date)

As for multiplayer..if it ends up happening...hopefully we get enough ppl with fast connections to host some interesting systems :-)
Vega Strike Lead Developer
http://vegastrike.sourceforge.net/
Guest

Post by Guest »

I think you guys have misread what I'm saying here.. maybe my fault partly for the wording:

1 ) GPL is not and cannot be violated by simply distributing pre-compiled binaries when all source is available. The source for these is always available, but the VS ' select distribution ' is a rolled distribution from the publicly available sources that is officially sanctioned. This binary has a known hash that is the first line of cheat prevention e.g it will put off many casual cheaters as they wont have the capability to reverse engineer it.

2 ) No mention has been made of charging for VS... thats not the point of this thread and not an area we should be clouding this theread with. This thread is supposedly about online development and ways to take that. Commercial exploitation of VS is a different arena.

3 ) There are so many ways to cheat ranging from code manipulation, to settings changes to use of bots (computer assisted targeting etc) with devices such as proxies to disguise this that the only way to throttle cheating is to be proactive. As for BZFlag, you only have to do a simple search on Google to see how many cheats there are for it... security by obscurity doesnt work: http://www.google.com/custom?cof=AH%3Ac ... &sa=Search

4 ) My thoughts reflected official master servers, not the ad-hoc arrangements that would occuer between user groups etc. You are always at liberty to play how you want and when you want in this environment. Im talking about a way to protect the game play environment for those users who wish that by a degree of policing. This is not a 'trojan' and to suggest so is somehwat naive to say the least.

Remember YOU ALWAYS HAVE THE CHOICE IN VS TO ROLL YOUR OWN AND PLAY YOUR OWN GAME.

To suggest that one size fits all and what you shout for alone should be the choice of all is nothing short of dictatorship and is 100% in the face of OpenSource, GPL and all they stand for... they are not concepts of free software, they are concepts in freedom of software. The financial aspect is only one small part of that.

Think about that before suggesting that my comments have no place and yours alone do. What I'm suggesting above is a possible path forward, not the only path. Its eneveitable that divergences will occur... in a way they already have with mods such as Privateer and VegaTrek etc... evolution is about change and diversification.

For me, as a long-time gamer and user of both Free and commercial UNIX for 10 years I feel I have a valid thought to give. The reality is that connections and hardware are ever evolving and you cant build to accomodate all player types. Take me, I used to play online wtih a 14.k modem.. then I thought i was an LPB and a hald when I was playing quake at 28.8k... then 33.6 and finally 56 then 56 x2 / KFLEX. ISDN was a screamer then DSL and cable... been there done them all. The reality is that in the policed gaming environment some game slow downs for users is possible on a modem connection... but thats no reason to deny all other players such access to accomodate the modem only users. Modem users online are something of a dying breed due to game bandwidth consumptions increasing and phasing them out / upgrading them to broadband.

I would say that I'm more than happy to accept any pre-compiled game binary available from an opensource tree in order to make the game more stable and to remove some of the cheat factor. I can only speak for myself, but one of the biggest kill-joys is watching a player do something awesome then immediately woner if they uised a cheat to do that.
Fink
Just a tourist with a frag'd nav console
Just a tourist with a frag'd nav console
Posts: 3
Joined: Mon Dec 27, 2004 9:35 pm

Post by Fink »

[-------]
[-------]

Sorry above reply was me.. wasn't logged in :-(

[-------]
[-------]
scheherazade
Developer
Developer
Posts: 427
Joined: Thu Jan 09, 2003 6:03 am

Post by scheherazade »

cheating can be eliminated simply by having an account server and game moderators with the ability to penalize someone.

and a spec-mode, where you can spectate someone.

you can just see if they're cheating, and if they are, do something to them.

kill them
take their $
ban them for some period of time
etc...

word gets around who cheats.

and since you need to work up to some financial statis in VS, simply making a new account woulnd't be that hot of aproposal for anyone who wishes to hide from his previous transgressions.

-scheherazade
jackS
Minister of Information
Minister of Information
Posts: 1895
Joined: Fri Jan 31, 2003 9:40 pm
Location: The land of tenure (and diaper changes)

Post by jackS »

may just be a diction issue, but -
eliminating cheating and/or otherwise making it "impossible" to cheat are stronger goals than I think are practical to pursue. Applying techniques, both social and in implemented code, that serve to mitigate cheating to negligible levels has always seemed a much more pragmatic phrasing of what needs to be done.
zazen
Trader
Trader
Posts: 25
Joined: Mon Nov 29, 2004 5:24 pm
Location: Brisbane damnit Australia damnit
Contact:

Post by zazen »

First Point --> modifying the game for personal benefit/gains is against the spirit of the GPL.. it allows you the freedom to change and modify code simultaeously granting other people the right to benefit from the changes you made.. In other words, a cheater must make the cheat publicly available or the person breaks the GPL and thus the law.

Lets have some proof that cheating is possible...

Picture the server binary as the variable A
The client as variable B
The interaction between server and client as C
The general public license as GPL

Thus GPL(A + B) = C

you can change the grouping and function of A and B all you like from the perspective of A or (A,B,C) but the fact remains that B basically has an identical perspective as A. There is probably a more defined mathematical model here that deals with the GPL. Though the GPL was designed as a "CATCH ALL" solution to preserve the right of B to change A and A to change B thus allowing for many possible outcomes at C.

I figure the moral of the story is this.. As long as both A and B are functions of GPL then C is valid. As long as A functions in a way that doesn't provide an alternate C and still is a function of GPL then only people who have the original B will get the perspective of C.

Do you think it's proof that legally cheating can happen?
I think it's proof that cheating is very possible and the cheat have virtually any degree of complexity and difficulty of changing B is at least as hard to change as A or harder (since you still have to partially understand A to change B to get a desired C).

Do you think it's proof that we can still have the original C as long as A does the following?
It's good news for VegaStrike because the only way to ensure an original B is to examine C which happens to be the easiest possible outcome from the perspective of either A or C. Since we don't try to guess B, rather C is used to get the value of B. To get B you require that A + B? = C so B? = C - A etc..

was that too wacky? hope i didn't make an idiot of myself here :)
Black tea, with a bit of milk.. and don't steal the sugar please!!
Alaskan
Hunter
Hunter
Posts: 79
Joined: Fri Sep 10, 2004 12:16 pm
Location: Australia
Contact:

Post by Alaskan »

This is me just dropping in...

Only the LGPL says you should send any code modifications back. The GPL says you have to make any dirrivative works avaliable.

How would people cheat?

The server validates everything like what you have, so you can't just keep saying you have 50 missiles.

You can only shoot forward, an aimbot can't help you there, and turrets already have AI for them.

HellCat, you logged before I could ask...Can you get the other network dev to contact me? Or just PM me with his details.Maybe I should have PMed you, well, I've already written this. I'll do both.
2 133t for joo!!!
hellcatv
Developer
Developer
Posts: 3980
Joined: Fri Jan 03, 2003 4:53 am
Location: Stanford, CA
Contact:

Post by hellcatv »

alaskan: I sent you an e-mail with contact info
Vega Strike Lead Developer
http://vegastrike.sourceforge.net/
martin
Hunter
Hunter
Posts: 80
Joined: Fri Oct 31, 2003 3:17 pm
Location: Berlin, Germany

Post by martin »

a few suggestions:

a) cloaking (if we really need such a feature): if something is cloaked, don't publish the position to the other clients. Nobody can misuse data that he doesn't have. This also solves video driver hacks (in space there are no walls behind which you hide anyway).
[digression, why we don't need cloaking]
Actually I don't see the point of a cloaking device. If you want to hide, just hide in the vastness of space. Of course this implies that Radar is not infinitely powerful - limited range and you need to "sit" longer to detect far away things. This could all be handled by the server - only broadcasting positions of in-range objects (of course the client can and should extrapolate moving objects and memorize positions of static objects - I consider this a legitimate use, a feature every navigation computer would have anyway). In an asteroid field there would of course be no way to distinguish a hidden station from an asteroid - ideal location for hiding stations (OTOH you could look for radio emissions or just follow a ship).

In summary, for a space game let's use the specifics of space and not force traditional concepts into the game. That way it will also feel a lot more like a space game and not like a flight simulator minus ground.
[/]

b) economics should be easy to check and compute on the server, as this is not as time critical as combat. Can probably also be done on a seperate machine to reduce load on main server. Also battle computations of different solar systems can be done independently and thus split on different servers, if necessary.

c) somehow I don't like the idea of signed binary executables, because I like to compile with optimizations for my CPU.

d) clients making the consistency checks and reporting cheaters: dangerous. I already see "applications" of this feature which allow to kick any player from the game by accusing him of cheating.
jackS
Minister of Information
Minister of Information
Posts: 1895
Joined: Fri Jan 31, 2003 9:40 pm
Location: The land of tenure (and diaper changes)

Post by jackS »

martin wrote:a few suggestions:

a) cloaking (if we really need such a feature): if something is cloaked, don't publish the position to the other clients. Nobody can misuse data that he doesn't have. This also solves video driver hacks (in space there are no walls behind which you hide anyway).
[digression, why we don't need cloaking]
Actually I don't see the point of a cloaking device. If you want to hide, just hide in the vastness of space. Of course this implies that Radar is not infinitely powerful - limited range and you need to "sit" longer to detect far away things. This could all be handled by the server - only broadcasting positions of in-range objects (of course the client can and should extrapolate moving objects and memorize positions of static objects - I consider this a legitimate use, a feature every navigation computer would have anyway). In an asteroid field there would of course be no way to distinguish a hidden station from an asteroid - ideal location for hiding stations (OTOH you could look for radio emissions or just follow a ship).

In summary, for a space game let's use the specifics of space and not force traditional concepts into the game. That way it will also feel a lot more like a space game and not like a flight simulator minus ground.
Cloaking, as it's currently implemented, isn't what we're looking for. At some point, active vs. passive sensors will actually matter. At some point ECM will be able to confuse sensors as to the exact nature and position of your ship. At some point there will be range effects on the accuracy of sensors. At some point there may even be effects due to other objects like asteroids, dust clouds, etc, but that's algorithmically harder.

The eventual implementation of cloaking should look something like an active medium range sensor confusion approach, wherein the vessel does not drop off sensors, but rather becomes impossible to pin down beyond being within some region of space, although the size of the region would diminish somewhat with range (hence medium moreso than short or long range, as at long range, you'd be better off with the "run silent run deep" passive sensors and muted output approach - I intend cloaks to actually light up sensors like a christmas tree - and at short range the effectiveness is such that the benefits become outweighed by the limitations).

For multiplayer, the positions sent will be those reported by the sensors, not the position of the actual vessel, that is, randomly generated outside the client (server side) with some probability function based on the cloaked ship's position, and then sent to such clients as would be sensing the ship. When we implement sensor error, we'll also have to implement various "unknown/uncertain" types/tags for transmitting what ship class has actually been detected.

As such, I still see a usefullness for cloaking as an active sensor-countermeasure. Not for hiding, but for making the class, vector, and position of the cloaked ship very difficult to determine.
d) clients making the consistency checks and reporting cheaters: dangerous. I already see "applications" of this feature which allow to kick any player from the game by accusing him of cheating.
No one said anything about clients being allowed to kick people. Indeed, I haven't seen mention of any individual client even being allowed to accuse people of cheating. Client side consistency checking is best handled as a distributed systems problem where one assumes some degree of collusion and corruption, and chooses sets of randomly selected clients of such size as the overal voting returned from the set of clients is probabilistically correct. Even if the set votes in accusation of cheating, the actual determinations should still be made by trusted machines - a probabilistically correct accusation is still a very good predictor mechanism for determining which players the trusted machines should be spending more time watching/reviewing backlogs for inconsistencies.
Planets

Post by Planets »

No more real cloaking devices ?
:(

Could you make real cloaking devices uber rare and difficult to find instead, and add the new electronic cloak?
Planets

Post by Planets »

For thoses who hate the GPL:
Please go away.

If you want a closed space game go buy on.
Vegastrike is completly free (speech etc), if you don't like it go away.
zazen
Trader
Trader
Posts: 25
Joined: Mon Nov 29, 2004 5:24 pm
Location: Brisbane damnit Australia damnit
Contact:

Post by zazen »

Alaskan wrote:This is me just dropping in...

Only the LGPL says you should send any code modifications back. The GPL says you have to make any dirrivative works avaliable.

How would people cheat?

The server validates everything like what you have, so you can't just keep saying you have 50 missiles.

You can only shoot forward, an aimbot can't help you there, and turrets already have AI for them.

HellCat, you logged before I could ask...Can you get the other network dev to contact me? Or just PM me with his details.Maybe I should have PMed you, well, I've already written this. I'll do both.
Sorry, I find this post to be wrong. Please familiarize yourself with GNU licensing, see here
Black tea, with a bit of milk.. and don't steal the sugar please!!
Halleck
Elite
Elite
Posts: 1832
Joined: Sat Jan 15, 2005 10:21 pm
Location: State of Denial
Contact:

Post by Halleck »

martin wrote:Actually I don't see the point of a cloaking device. If you want to hide, just hide in the vastness of space. Of course this implies that Radar is not infinitely powerful - limited range and you need to "sit" longer to detect far away things.
...
In summary, for a space game let's use the specifics of space and not force traditional concepts into the game. That way it will also feel a lot more like a space game and not like a flight simulator minus ground.
Well, hiding is only one of the possible functions of a cloaking device. It could also be used to sneak past enemy fortifications (useful for blockade running), avoid cargo scans, and launch surprise attacks. None of these tasks can be accomplished by just sitting out in space.

Maybe we can have both optical cloaking devices and "stealth" armor in addition to sensor-confusing mechanisms. The optical cloaks could be balanced by losing their effectiveness based on activity; perhaps they can work a bit like shields in that certain activities drain them. These won't necessarily be the same activities (such as using the SPEC drive), but maybe things such as firing or using the afterburner.
Also, optical cloaks could produce distortion when a ship is moving- harder to see in space, but fairly easy to spot against a planet because of the way the light bends. I'm envisioning something not unlike the Starcraft cinematics, especially the one where the Protoss fleet passes by a planet and uncloaks.

Stealth armor/coating could be a space-age equivalent of the stuff they use to plate B-2 Bombers, which absorb and deflect radar. Certain types scanners would not pick up the ship, and certain types of armor would work against different types of scanners.
Post Reply