Slashdot Mirror
Open Source FPS Blood Frontier Releases Beta 2
An anonymous reader writes "The open source FPS Blood Frontier has now made their beta2 release. From the article: 'After many months of development, and massive amounts of input from the public, we are proud to present you with the new release of Blood Frontier, v0.85 (Beta 2). This new version totally redefines and improves the game in many ways, creating a whole new style that makes it almost nothing like its predecessor.'"
Judging this book by its cover, this looks like a remake of what I used to play more than a decade ago with Quake 3: Team Arena. I watched the video, looked at the website, but all of the improvements listed there boil down to either move better, or kill more stuff. Anything actually original about the gameplay that makes it Killzone 2 kind of fun?
Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
"proper anticheat system" = server admin who bans cheaters. problem solved.
So not really a beta then if it's been changed so much that it's nothing like the 1st beta release.
More like an alpha release and they're still sorting out the requirements.
Bad analogies are like waxing a monkey with a rainbow.
It looks interesting from the video, but I'm tired of playing Quakelikes after all these years. And I only played Quake for the first time about five years ago.
Clones and similar games are inevitable, of course. I just wish someone would start cloning other hallmark FPS games, like Serious Sam, Doom, Perfect Dark, etc. I'd really like to see some sort of espionage-based FPS out there in open source; something like Splinter Cell, Perfect Dark, or Rainbow Six. Fun things like cloaking devices, remote cameras, etc.
But since Christmahanukwanzaa is coming up, I'd like to ask Santa for a true-to-the-spirit-of-the-original TRIBES remake. I still play that game nowadays but it isn't as much fun without new blood - and the game makes it very difficult to get new players to come in (thanks to the multiple barriers to entry such as a ridiculous number of mods, custom patches, unofficial master server (because the original was shut down), etc.
Random Thoughts From A Diseased Mind (Not For Dummies)
Another unoriginal arena shooter. Sure it looks very pretty but it doesn't do anything that hasn't already been done a million times before.
Please stop making these now.
Summation 2
Any sufficiently advanced cheating is indistinguishable from skill.
One that hath name thou can not otter
It still ruins it kinda. Sure you get off the idiots who run around at 100x speed aimbotting and killing everyone, but discreet use of esp hack that shows you where enemies are or if someone is coming behind you will still give major advantage to the cheater. It will not ruin the game completely, but it still makes it kind of stupid.
Actually this is and will always be a major problem with open source online games. You don't even need to debug assembly and create a hack for it, you just need to edit the source code and build your own client. Open source anti-cheating system has the same problems too, and in addition open source community would probably yell against locking down the client with such system. Sure, server admins can still ban the obvious cheaters, but this is one of the things commercial games (and commercial anti-cheating software) will always have advantage over open source games, at least until we can actually just render the player screen on server and transmit it over the internet.
Actually this is and will always be a major problem with open source online games. You don't even need to debug assembly and create a hack for it, you just need to edit the source code and build your own client.
Not always.
It can be solved by not sending to your client any information you shouldn't know as player or character.
It's fricking hard, but not impossible.
Isn't that on similar level of "hard" (perhaps even on greater...) to simply streaming the image to client with sufficiently low latency?
One that hath name thou can not otter
Not even close. As long as computers are better at tasks than humans, the issue remains.
i.e. Think about an aim-bot. A computer is going to be more accurate and faster than a person with the same data. And even if you do the whole streaming-game thing, bot likely will still exist, that process the stream and emulate user actions far quicker than a human could
Feel free to mod this post as flamebait, but I feel it's time to rant about the Open Source gaming community. It seems to me whenever there's a new Open Source FPS that comes out, it's just another pathetic Quake clone. Sure the trailer videos *look* cool, and sure the screenshots are rendered at high resolution, with all the bells and whistles enabled. All is good until it comes to the actual gameplay. It's disappointing when all the freetards (excuse my french) drool over another cheezy clone (merely because it's Open Source, but not of it's merits alone) that's no different from the previous hundred clones that came before it. Boring and unoriginal.
Which brings me to my point: WHAT ABOUT WARSOW?! This game has been out for years, it's free, the source code is GPL'd, runs on windows/linux/mac, and above all the gameplay takes the Quake shooters to a whole new level. In all of my 15+ years of gaming, warsow is by far the most complex and elegant FPS to date. Imagine playing quake2/quake3, now imagine that on crack. That is warsow. It's not another lame re-skinned quake clone like it's predecessors. A quick search on slashdot shows only one post referencing it... ONE POST!!
The community is small and has been diminishing over the past couple of years. Which is quite surprising for a game with such immense potential. My only guess is this: the game is too hard. Yes I will admit that the learning curve is steep, but that's half the fun right there! You would think a community of opensource folks (who love to tinker with their own systems, to learn and read and gain a better knowledge of the inner-workings of their respective systems) would be chomping at the bit to take on a game that requires some sort of learning. If you're willing to spend 5+ hours trying to decipher an archaic perl regex, you shouldn't break a sweat trying to learn how to rocket jump over the period of a half an hour or so.
You would think a community that looks down on proprietary cookie-cutter products would embrace originality and innovation in their games, but it's starting to look like the Open Source gamers are painfully similar to their proprietary counter-parts. Same cookie-cutter crap as before, only difference being the price of their engine. </rant>
A computer is going to be more accurate and faster than a person with the same data.
If you've actually got the algorithms to back that up, you'd be a pioneer in AI.
Want to prove it? Dial Goog-411. That's right, that's Google doing voice-recognition. And their massive cluster still sometimes has to make you wait a few seconds while they try to figure out what the hell you said.
So, getting back to what you said:
Think about an aim-bot.
There are generally going to be two things happening here:
Either you've discovered a surefire way to distinguish real targets from noise, in which case, the army probably wants to talk to you... ...or your aimbot is going to both be too perfect at headshotting people, and it'll make stupid mistakes that only a bot would. This is where it becomes painfully obvious to a good admin and/or good server-side passive anti-cheat what's going on.
I'd also suggest you look into the Robo-Olympics, and various programming competitions, and the general failure of AI in single-player games. Again, it is in general either not possible to do better than a human, or it's so obviously better than a human that you get banned.
Don't thank God, thank a doctor!
From the wiki:
The game is a single-player and multi-player first-person shooter, built as a total conversion of Cube Engine 2
That alone adds a dimension that simply won't be there in Quake 3: real-time, multiplayer map editing, on the server, while others are still shooting each other.
Don't thank God, thank a doctor!
Looks like a low-quality Q3A remake.
I don’t think it can beat Q3A CPMA (and Defrag) though. Especially not with the XreaL engine and High Quality Quake models and textures.
Any sufficiently advanced intelligence is indistinguishable from stupidity.
Actually this is and will always be a major problem with open source online games.
Actually, for some genres, an authentication scheme based on PGP's web of trust could be used to restrict who is able to participate on a match. Anybody that is obviously too good to be true could be collectively banned. Those who can't be subjectively detected, don't hinder the gameplay.
If I had the time, I would implement something like that for Frozen Bubble... (oh, but I *hate* Perl)
`echo $[0x853204FA81]|tr 0-9 ionbsdeaml`@gmail.com
It's not really hard to make aimbot work so that it's not so obvious, it's only bad cheats that just jump to point to enemys head and shoot. It's obvious of course. But it's trivial to add smoothness and human-like errors to the movement. The data is there. Client knows where enemy is (because it has to be drawn). It's a matter of few cpu cycles to calculate the new direction to aim at and make command to shoot. No human can possible, ever, work that fast.
Another problem is ESP, warnings and other such cheats that do not directly interface gameplay, but give cheater a huge advantage because he gets a lot more info than other players. Then it's just up to him to act in a way that isn't so obvious.
Only the stupid cheaters are caught by server admins. And the actual non-cheating, but good players get kicked by trigger happy admins. There IS need for anti-cheating software.
Is there no such thing as sanity checking for this kind of thing? "Player has shot the last five enemies in the face at exactly the same level below the (team coloured?) helmet. He's probably got an aimbot running." or "Player was in this room, he's now in the courtyard, why the hell wasn't he in the corridor before that? Wallhax."
Finally had enough. Come see us over at https://soylentnews.org/
Anybody that is obviously too good to be true could be collectively banned. [...] If I had the time, I would implement something like that for Frozen Bubble
So you like puzzle games, and you want to ban players who appear too good to be true. Would you end up banning Jin8 from playing Tetris? He's actually that good. (Fast forward to 5:00 and watch for 20 seconds and be amazed.)
If you have too many G rating you will of course not be allowed on most servers.
But if an experienced player of previous games in the same genre ends up randomly matched with less skilled players due to too few games played to establish an Elo ranking, expect a significant number of spurious Gs due to the vast skill differentials. Let's say Jin8, one of the half-dozen players worldwide who have achieved the "Grand Master" grade on Tetris The Grand Master 3: Terror-Instinct, has just signed up on your puzzle game server. He could play falling block games almost with his eyes closed, as shown in the famous video starting at 5:10. Blink, one of the respected members of Hard Drop Forum, isn't quite as skilled as Jin8, but he can still clear 40 lines in a hardcore Tetris clone in under 30 seconds (compare my 60 second time and casual players' 120 second times), and even videos of him "sucking" at a casual Tetris game are probably far better than you could do. Would they be a 3 or a G?
And some servers will require that you have a certain consistency to be allowed to play.
Then how does a new player earn consistency if very few servers allow players who haven't already earned consistency to play?
good content is hard to get for free.
This isn't the case for computer programs, as shown by the free software movement over the past two decades. So why is it the case for meshes, textures, maps, audio, and scripts?
This is not an open source project. From data/textures/readme.txt:
(C) 2007-2009 Blood Frontier Team, all rights reserved.
The "textures" package included in Blood Frontier may only be distributed
with the Blood Frontier package. Redistribution or repacking outside
this context without the author's consent is strictly prohibited.
If you want a real open-source shooter that rocks, try Nexuiz.
OS Reviews: Free and Open Source Software
A computer is going to be more accurate and faster than a person with the same data.
If you've actually got the algorithms to back that up, you'd be a pioneer in AI.
Want to prove it? Dial Goog-411. That's right, that's Google doing voice-recognition. And their massive cluster still sometimes has to make you wait a few seconds while they try to figure out what the hell you said.
I would think that there's a slight difference between determining the vector from one known object (the local player) to another known object (an enemy in the local player's viewport) and doing voice recognition.
Because Creative/Artist types want to get paid, good reliable artists are hard to find.
Programmers also want to get paid, yet many work on free software for free in their spare time, and some even get paid by big companies to develop free software like Linux, Qt, Firefox, and OpenOffice.org. I'm still not seeing the cause of the difference between the executable and nonexecutable portions of a game here.
... I'm always curious why the open source FPS games look like they are about 7 or 8 years behind the closed-source industry.
Its like they're not even trying to compete. Go to the Game Developers Conference, guys... Take some notes... See what the top devs are doing in the future and start doing that NOW. Then you'll catch the wave at the right point.
open source FPS always catch my attention long enough to notice this consistent failure to get with the times.
Ok, I hope not to be modded troll for this, but probably will be.
I have played a lot of FP shooters. The innovation from one generation to the next in terms of graphics and stability has been wonderful, brilliant, and lacking in magic.
What is missing, and what could make the next big FPS is gameplay. Anyone who wants to do it right needs to sit down, play with 3-5 friends some Renier Knizia board games
http://www.boardgamegeek.com/boardgamedesigner/2/reiner-knizia
as an education is what gameplay could be, compared to what it is. Don't call me a crank if you have not played Modern Art, Tigris and Euphrates, and Through the Desert.
it's trivial to add smoothness and human-like errors to the movement.
At which point, it just becomes an arms race. Add a human-like delay, and a human can win. Add "human-like" misses, and chances are, they'll show a predictable pattern.
Client knows where enemy is (because it has to be drawn).
There are several ways to fool this kind of cheat, however. One possibility (making it up as I go) is to send false information about an enemy position, such that it likely won't be drawn (behind/in a wall, or with an invisible model) and watch for clients that immediately snap towards that target.
Another problem is ESP, warnings and other such cheats that do not directly interface gameplay, but give cheater a huge advantage because he gets a lot more info than other players.
To a certain extent, you can't really do anything about this. For example, if your game is something like Doom 3, so you've got a lot of darkness, or maybe you've got a nice sniper spot that's hidden by HDR bloom, you might have a player who is just very observant, or you might have a player with worse hardware so they have to turn effects off, or you might have a player who deliberately turns those effects off (and maybe loads bright orange models for the other players) in order to see.
On the other hand, aside from the misdirection I've mentioned, one thing that helps a lot is to not send information the player doesn't need. For example, Counter-Strike traditionally sends positional information to all players, all the time. One of the many admin mods only sends updates for players you'd be able to see/hear.
Don't thank God, thank a doctor!
Since there's no "-50: Hopelessly Wrong", I'll sacrifice modding parent as Overrated to post a reply that will hopefully re-clue anyone who reads and believes it.
Even with access only to the data you "should know", it's still TRIVIAL to mod a client in ways that provide significant advantages. No offense, but parent has absolutely no idea what he's talking about, and obviously no actual experience in this area.
Rather than listing 20 or 30 trivial cases that disprove your claim, I'll just take the most obvious one.
An enemy is clearly visible a hundred yards away. I think we can all agree that that's "information you should have as a player", right? My client knows where he is, because it has to draw him. So, with some trivial math, my client is capable of instantly targeting him and shooting for me.
You say it's not impossible to stop cheating, just "hard". Start with that one, then we'll move on to the more complex cases...
(And no, even just streaming the game doesn't in any way resolve this, even if it wasn't impractical. Trivial image analysis will pick out the enemy player by motion, color, etc)
If you've actually got the algorithms to back that up, you'd be a pioneer in AI.
You're joking, right? The theory behind an aimbot is pretty much as simple as it gets. And it's not like you have to kill every person on the map, you could easily use the aimbot to achieve "snapping" to the target or what not.
Want to prove it? Dial Goog-411. That's right, that's Google doing voice-recognition.
While I'm sure you've proved something, it has nothing to do with this. If the task isn't suitable for a bot, don't do it with a bot. You'll notice I also qualified it with a statement "as long as a computer is better at the task" part. Things like social element and group work is going to be too difficult.
your aimbot is going to both be too perfect at headshotting people, and it'll make stupid mistakes that only a bot would
It's a silly point, as it's all down to implementation of the bot. The bot doesn't need to do more than just assist you to be a huge advantage.
One such bot that comes to mind is: tibiabot It only uses data accessible to a person, and only does actions a person could do. And it primarily sits on the background until needed. An example of its use is to automatically hit someone when they're on low health (before they have time to heal) or to automatically heal (with an optional rand(x,y)) delay on low health. Or you can use "combo" features where multiple people playing will all target the same person automatically etc. And even to automatically heal a friend, something that would normally go missed because a player is too focused on something else. It can make a beginner player far more effective player than an advanced player in battles. And is all done without extra data.
Strangely we see all kinds of abuse in Spring but noone seems to bother altering the engine itself to make a map hack.
Justice is the sheep getting arrested while an impartial judge declares the vote void.
An aimbot that works from only an image would not be that easy, true but an aimbot that interfaces with the game state and simply reads player positions from there is trivial since AI agents already do that (of course usually with some artificial inaccuracy and delay to prevent them from simply wiping players out).
Justice is the sheep getting arrested while an impartial judge declares the vote void.
This won't help too much. There will always be ways to cheat. Especially if you know exactly how the server is checking the validity of the client.
The only solution is for the server to check if the client uses an unmodified version of the game. Exactly that is, what "trusted computing" is promising to deliver.
I still don't believe trusted computing is possible if the client has total control over his machine (is root) but at least in theory that's the only answer I can think of to solve the cheating problem.