Siiiigh.. it's supposed to be deleting the slime on collision.. instead it pushes it... I don't even know how this is possible...
[vid]http://anotherprophecy.com/webm/sigh.webm[/vid]
[QUOTE=false prophet;48855773]Are you doing to read it to them or rely on them to read it?[/QUOTE]
I haven't decided yet. It depends on exactly how high my friend is, and if he's still able to read.
I am not an expert on the effects of LSD.
Okay, here's what you need to do. Two different stories, one read to the victim, the other let the victim read him/herself.
Read it first, it will be hilarious. Then let him read a different rant story himself, out loud for even more hilarity clinton.
[QUOTE=false prophet;48856112]Okay, here's what you need to do. Two different stories, one read to the victim, the other let the victim read him/herself.[/QUOTE]
Ooh, I like that. I'll see about writing a pattern during class to generate them with. Otherwise I can just pull a couple of the old classics off the ancient Moist Banana blog.
Quantum AI doing a 4x4 sudoku game (unrelated to neural network project):
[img]http://i.imgur.com/WhmEhJx.png[/img]
[img]http://i.imgur.com/vUV6KGY.png[/img]
[url]http://hastebin.com/megifunaci.java[/url]
[editline]8th October 2015[/editline]
Quantum AI is a very fancy name for an AI that cycles through possibilities before picking the best one.
[QUOTE=cra0kalo;48855195]Well you just don't do what VAC scans for pretty straight forward lol.
[editline]8th October 2015[/editline]
Oh yeah I forgot this was facepunch woops. Another retard which thinks a game ban is vac rofl. For your information that isn't even csgo related. Open your eyes fool
[url]http://steamcommunity.com/sharedfiles/filedetails/?id=531533701[/url]
wait NO i got an idea. It's an [B]Overwatch bypass[/B] yeah thats it.. thats something your brain can probably comprehend lol.
[highlight](User was banned for this post ("Flaming" - SteveUK))[/highlight][/QUOTE]
he was making a joke but apparently thats not something that your brain can comprehend
[QUOTE=Ott;48856628]
Quantum AI is a very fancy name for an AI that cycles through possibilities before picking the best one.[/QUOTE]So in essence it brute forces? Or is there more to it?
Does anyone here have experience with game balance (particularly for an ARPG)? Or does anyone know a good place for discussion on topics like these? I wrote something about how I'm thinking of approaching the way to create the stat systems and how to balance them in my game and I was wondering where I could get some feedback. Also, here's what I wrote if you wanna give feedback on it (what's wrong with it, why it won't work, etc):
My game is going to be kinda like an ARPG so assume I'm talking about this for now (Diablo, Path of Exile, and such)
In those games, behind all the stats and items and shit, all it really comes down to is how much DPS you have and what's your chance of dying with the current playstyle, let's call this ROD (Risk of Death). So from a design perspective you have these two numbers to work from, DPS and ROD. A playstyle with high DPS and high ROD would be a very squishy melee character, a low DPS and low ROD one would be a very tanky ranged character.
DPS is easy to model. You just figure out how many hits you want it to take to kill enemies at certain levels (and level here can mean an actual level or just at a certain progression level that the character is in, i.e. in Dark Souls this can be "after you've gotten this many thousand souls") and then from there you can add all your offensive stats and make sure that when it all adds up, the DPS generated from them should remain in the "zone" you want player's DPSs to be in at that level, like, if you have 1000 [resource] your DPS shouldn't be lower than 100 but also not higher than 300, and you just architect things so that this is true and that it also accommodates for various playstyles. Also, unlike in the CoD example you gave, in an ARPG, DPS doesn't really take into account player skill that much. Yes, in a shooter it's a lot harder to balance this because player skill is encoded in the offensive stats, but in an ARPG it's the opposite, assuming similar stats, every player will play the game optimally offensively, what matters is how they play the game defensively, i.e. do they kite enemies? Do they heal as often? Did they bring the right kind of potions to deal with status effects? And so on. So this is why DPS is easy to model in this case (unlike with CoD).
OK, so DPS is easy to model but now you have to also model the actually hard part which deals with player skill. The way I'm imagining of doing it (and I don't know if it works) is to assign a value to various different playstyles and this value will correspond to the character's current Risk of Death.
If you're really a really squishy melee character with low DPS, either because you suck or because you're investing points in non-combat related stats, the more you go on in the game the higher your ROD should get, because enemies get harder, they hit for more and they have higher HP, so you're just spending more time overall on each battle. So from a design perspective this works kinda similarly to DPS, you want to make it so that at a certain level, like, the player has 1000 [resource], you want the available ROD zone to not be lower than min_value and not higher than max_value. If it's lower than min_value then the player will eventually get killed (and why it's lower is either because he doesn't have the appropriate stats or he sucks at the game), if it's higher than max_value then the player will never die and the game will be boring. So you have to architect things in such a way that it's really hard to get lower than min_value unless the player is a complete retard, but also that it's really hard to get higher than max_value. Also the final ROD value, unlike with DPS which is just a simple sum of all offensive stats, is the sum of all defensive stats times some factor K. This factor K should be what accounts for different playstyles and different skill levels, and this is the hard part to get right.
Let's say we have 5 different basic playstyles: melee tanky low DPS, melee squishy with high DPS, ranged no kite, ranged kite and ranged kite with dodge. We want a way to encode those different playstyles into a value K that will multiply the current sum of all defensive stats and that will make that value (sum of all defensive stats * K) fall within the "zone" we want this value to be in so that players aren't too challenged nor too bored with the game.
Let's say for now that K ranges from 0 to 1, 0 being high ROD and 1 being low ROD. Melee tanky has lower ROD than melee squishy with DPS, probably. If the melee DPS is so high that it can kill enemies before they even hit you then this doesn't apply, otherwise it seems reasonable. All ranged in general have lower ROD than melee no matter what, and a ranged player that doesn't kite enemies will probably get hit more, so he has higher ROD than those who do kite, and between those who do kite someone who kites using the dodge skill (which lets you cancel the current attack animation to move) will probably have even lower ROD than someone who doesn't. So in our 5 playstyle example we have the following ROD setup when compared to one another: melee squishy < melee tanky < ranged no kite < ranged kite < ranged kite with dodge. Then we could assign numbers like 0.1 0.2 0.5 0.7 0.8 but right now it doesn't make sense because we don't know what these numbers should look like exactly, but we do know how they related to each other and from this we can sort of model different playstyles and different skill levels. It's also important to note that the calculation to come up with this value takes into account DPS. For instance, melee tanky has lower ROD than melee squishy, but if melee squishy's DPS is really really high then battles last for a really small time and maybe his ROD gets to be lower than a melee tanky character.
Does this make sense?
i'd just try and math it out roughly then let lots of people with different experience levels with rpgs test it out.
I was just on a lecture where King.com (candy crush saga makers) held a little talk about what King.com is and how they develop games. Apparantly they've just started using this new game engine Defold which looks epic. You code in lua and can build for all major platforms, with hot reloading code over Wi-Fi for phones and all cool shit.
Anybody heard of it before? It's in beta currently and I managed to get a key!
[URL="http://defold.com"]defold.com[/URL]
I feel like most of the people they want to apply to that either already have pathfinding experience, or are the types of people who will just look up and implement a pathfinding algorithm to solve it.
[QUOTE=geel9;48857965]I feel like most of the people they want to apply to that either already have pathfinding experience, or are the types of people who will just look up and implement a pathfinding algorithm to solve it.[/QUOTE]
I think the point of the challenge is not to make your average verbose pathfinding A* implementation but to really impress
When I think about it now, A* doesn't seem that hard really (never implemented it though). If you can traverse nodes with coordinates and make heuristic function, then you are pretty much set I think.
[QUOTE=itisjuly;48856881]So in essence it brute forces? Or is there more to it?[/QUOTE]
It cycles through each cell, checks if ot's zero, makes a copy of the game state, and cycles through numbers until the shard has less mistakes than the original game state.
[QUOTE=Donkie;48857770]I was just on a lecture where King.com (candy crush saga makers) held a little talk about what King.com is and how they develop games. Apparantly they've just started using this new game engine Defold which looks epic. You code in lua and can build for all major platforms, with hot reloading code over Wi-Fi for phones and all cool shit.
Anybody heard of it before? It's in beta currently and I managed to get a key!
[URL="http://defold.com"]defold.com[/URL][/QUOTE]
Oh boy...I really hope to get an invitation
[QUOTE=geel9;48857965]I feel like most of the people they want to apply to that either already have pathfinding experience, or are the types of people who will just look up and implement a pathfinding algorithm to solve it.[/QUOTE]
How about you try to hire some programmers first, and see how that goes.
Its there to root out retards, its not the deciding factor to get the job.
Taking everyone who can somewhat solve that will already have you down to the 5%.
[editline]8th October 2015[/editline]
[QUOTE=adnzzzzZ;48856996]wall-o-stuff
Does this make sense?[/QUOTE]
Its pretty hard to model game mechanics, especially in an ARPG, where you explicitly trying to have systems as complex as possible for the sake of character building, and trying your best to avoid having one obviously best build, while still making sure every build is meaningfully different.
Create some base guidelines with a small set of items, test those, balance them, and then align new items relative to them.
There are people who have tried applying models to "simple" card games, but once you get into an actual game and realize that a 8% range difference, or a 0.25 animation on a spell, can make all the difference in the world its going to go down the shitter real quick.
[QUOTE=Trebgarta;48857774][url]http://career.paradoxplaza.com/jobs/5763-are-your-programming-skills-up-to-paradox-standards[/url]
This seemed interesting to me, thought maybe people here might give it a try. I would, but I didn't because the big green button says "APPLY" and I don't have any pathfinding experience anyway.[/QUOTE]
I'm glad they specifically say what they want you to write the solution in. I'm so sick of "any language you'd like" bullshit. Firstly, it's not even close to any language because it almost always means C(++), Java or Python. Secondly, you're not going to work in any language you want if you do get the job. Why even offer it?
"Any language you want, as long as it differs from C only in the placement of the curly brackets"
[QUOTE=Darwin226;48859205]I'm glad they specifically say what they want you to write the solution in. I'm so sick of "any language you'd like" bullshit. Firstly, it's not even close to any language because it almost always means C(++), Java or Python. Secondly, you're not going to work in any language you want if you do get the job. Why even offer it?
"Any language you want, as long as it differs from C only in the placement of the curly brackets"[/QUOTE]
I disagree, languages/software comes and goes and are just a vehicle for problem solving. If you can solve it in one language or environment, you can solve it in another. Why lock people into something that is effectively meaningless? Assuming you get the job they're going to teach you their current pipeline and environment because that's how developing on a team works, but that shouldn't be a bar for getting your foot in the door.
I know from personal experience that as long as you're not solving it in something intentionally obscure and they say any language, they truly do not give a fuck what language you use.
[QUOTE=Cold;48858924]
Its pretty hard to model game mechanics, especially in an ARPG, where you explicitly trying to have systems as complex as possible for the sake of character building, and trying your best to avoid having one obviously best build, while still making sure every build is meaningfully different.
Create some base guidelines with a small set of items, test those, balance them, and then align new items relative to them.
There are people who have tried applying models to "simple" card games, but once you get into an actual game and realize that a 8% range difference, or a 0.25 animation on a spell, can make all the difference in the world its going to go down the shitter real quick.[/QUOTE]
I don't feel comfortable not having some kind of framework to guide me. Either way, the complexity of the system doesn't really matter because in the end it all boils down to two values that should remain within certain ranges (which I'm planning to make wide enough) and not be exactly this or that.
Like, I want the viable stat progression for level 10 to be: it takes 2-5 hits to kill a certain enemy while having a 0.3-0.8 risk of death value. Some builds and stat combinations will give me 2 hits and 0.3 ROD, others will give me 5 hits and 0.8 ROD, others will go out of this range and they'll either be too OP or too UP, all I have to do is make it so that all paths available for progress don't make the values go outside the range.
I feel like there can be some problems of circular dependencies where I change 8% of something here and it has a cascading effect that changes a lot of other stuff, but I think that the system complexity, various items and resources will come from solving these issues instead of starting from the items and then trying to balance them by brute force. If I want a certain playstyle that needs the 8% of X here, but this 8% makes it too powerful because of the cascading effect there in Y and Z, now suddenly I can create a system mechanic that everytime I get some improvement on X I also get punishment on Y and Z, like, say, items with increased attack speed have lower damage/AoE increases or even have damage/AoE decreases in them. So in my game this will be a fact of life, +APSD = -DMG and -AOE. And you repeat this idea of fixing these issues by introducing systems long enough and you have a sort of organic complexity generated out of necessity and that is internally consistent, instead of you just setting some random guidelines and bruteforcing from there.
[QUOTE=FlameCow;48859845]Finally got around to implementing proper knock-back in my game.
[vid]https://my.mixtape.moe/xejeaa.webm[/vid][/QUOTE]
From what I can see the attacks feel like they're fighting in molasses
[QUOTE=Darwin226;48859205]I'm glad they specifically say what they want you to write the solution in. I'm so sick of "any language you'd like" bullshit. Firstly, it's not even close to any language because it almost always means C(++), Java or Python. Secondly, you're not going to work in any language you want if you do get the job. Why even offer it?
"Any language you want, as long as it differs from C only in the placement of the curly brackets"[/QUOTE]
While I agree, I think it's nice to get the choice. If the fact that if the language that you did choose to write it in is seen as contrarian to them, it's probably a sign where you wouldn't like to work anyways. :v:
(ie, if they said that and I wrote it in Racket or something else I'm confident working in, and they'd be all "oh no the parenthesis!", I could probably conclude that they're shit.)
I think it's still a good thing to do (because why would you want to waste time writing in something you don't want, for something which might not even grant you a job?), in a sign of good faith [I]even[/I] if you know you'll be writing in another language when you actually work there.
It'd be okay for me to write something I wouldn't necessarily pick as a first choice on the job, but for a job interview which guarantees me nothing, I want the choice.
[QUOTE=Socram;48859609]I disagree, languages/software comes and goes and are just a vehicle for problem solving. If you can solve it in one language or environment, you can solve it in another. Why lock people into something that is effectively meaningless? Assuming you get the job they're going to teach you their current pipeline and environment because that's how developing on a team works, but that shouldn't be a bar for getting your foot in the door.
I know from personal experience that as long as you're not solving it in something intentionally obscure and they say any language, they truly do not give a fuck what language you use.[/QUOTE]
Does Haskell count as intentionally obscure? How about some Lisp? What if you solve it in assembly because you like hurting yourself?
I truly doubt you'll find many companies where you can literally submit your solution in any of these languages and they'll be able to read it.
Of course, this is all based on my statistically irrelevant sample size but I did have a Google recruiter tell me exactly what I wrote above. "Any language...", then when asked, "...that's Java, C++ or Python".
[img]http://i.imgur.com/uKGBmNL.gif[/img]
[QUOTE=AntonioR;48860298][img]http://i.imgur.com/uKGBmNL.gif[/img][/QUOTE]
I said "hooray" out loud the second time I watched him jump. It was great.
Finished my knife!
[IMG]http://i.imgur.com/we9n4WJ.png[/IMG]
It could be better, but hey, it was my first textured model. :v:
It'll do as a placeholder for now, so I'm happy with it.
So playing Warhammer 40k in Tabletop Simulator was a thing me and a couple of mates saw potential for as soon as first seeing the game spring up, but the game's functionality at the time wasn't conducive to playing just yet.
Skip to early this year, TTS' interface and feature set was right spot on for playing 40k, so now it was just the models. A little digging into extracting DoW models eventually uncovered the old tools and scripts used by modders (still active to this day!), and pretty soon using a clunky method;
[I]- Mate would use his copy of 3ds Max to extract a model and animations, export to FBX
- I'd import this into Blender, have to fix bone skinweighting (issue with Santos' maxscript)
- Scale, pose and stick on top of a correct base
- Manually combine the textures in GIMP into a single map, change UV's in Blender to match (easily the most time consuming part)
- Export final OBJ for use in TTS[/I]
Nice, and way faster (and cheaper!) than putting together a real model, but I saw the manual UV editing task as quite repetitive, and was surprised there wasn't a tool built into Blender for combining textures and UV's automatically. So I put together a quick Python script which takes an OBJ and its textures, slaps them together end to end, and updates its UV coordinates to match, removing that easily hour long job entirely.
Skip forward a bit further, my mates and I am playing a lot of 40k games amongst ourselves, each with an army or two ([URL]https://imgur.com/a/q2mYm[/URL]).
[t]https://i.imgur.com/jxqhHB1.jpg[/t]
But whenever someone else wants to join in, we've got to go through this entire process either for them, or teach them how to do it, and it isn't exactly straightforward.
So on and off during the last couple of months, I've been juggling a couple of projects, one of which is an improved texture packing tool which crops each texture to just what is used before packing, to save bringing in entire maps just for a single gun texture, for example (kitbashing has this scenario a lot).
I've also been working with lots of help from a mate on an addon for Blender which imports Relic Dawn of War WHM files straight, no need for Santos' old Maxscript (or 3ds Max, for that matter).
[vid]https://my.mixtape.moe/rkgpny.webm[/vid]
So far, the script supports practically everything needed -- reading in meshes, textures, bones and animations.
I've still yet to nail converting the animation bone's space into Blender's (it differs from the initial skeleton definition!), leading to some interesting results:
[vid]https://zippy.gfycat.com/TalkativeDamagedChamois.webm[/vid] [vid]https://zippy.gfycat.com/IcyVerifiableCaimanlizard.webm[/vid]
Surprisingly, even using it as the sole reference, somehow our implementation of reading in the bone skin weights doesn't suffer from the bug which plagued Santos'.
Really fun little project, I've been learning a lot, and playing with Blender's Python API is surprisingly lots of fun and fairly straight forward.
Once I've finished up tying the UI and squashed the last breaking bugs, this addon combined with my finished image packer script will make it very, very simple for people to extract models from their own Dawn of War game and bring them right into TTS to play games like Warhammer 40k with. And completely with free tools (and no slow, clunky 3ds Max 2010 + scripts).
A sudoku solver [url]https://gist.github.com/LukaHorvat/73669d989e9021ed25ee[/url]
I can't believe how short it turned out. The actual algorithm is the last 10 lines.
It solves any sudoku. Even ones that have multiple solutions. Finds all of them.
You can also parallelize it very easily. It's just `concat $ parMap rpar` instead of `concatMap` on the last line.
[QUOTE=Donkie;48857770]I was just on a lecture where King.com (candy crush saga makers) held a little talk about what King.com is and how they develop games. Apparantly they've just started using this new game engine Defold which looks epic. You code in lua and can build for all major platforms, with hot reloading code over Wi-Fi for phones and all cool shit.
Anybody heard of it before? It's in beta currently and I managed to get a key!
[URL="http://defold.com"]defold.com[/URL][/QUOTE]
As much as I love Lua, I'd rather make something from scratch than be associated with King.. and they're not using LuaJIT :(
WAYWO, I need some advice! I made this alt account just in case.
So I recently started a new job as a graduate software engineer. The company sounded great, and we were promised programming training, hands-on experience, all that good stuff. They take on people of all skill levels, provided they have SOME prior knowledge of programming. Anyway, I got there earlier this week and everything seemed great – the people are really nice, and the company isn't too big that you never see the CEO, but not too small that you know everyone.
Anyway, a few days later and we were assigned our desks, and were called in to chat about what we'd be working on first. The project sounds awesome – I can't say much about it, but it's very interesting. However, things took a sharp turn for the worst when we were told what part we'd e playing in this. We are the testers. OK, not so bad right, at least I can write some unit tests? Wrong type of testing! The type of testing I'll be doing is literally just USING the software to make sure it works correctly. The client has provided requirements for what the software should do, and it's my job to operate the software in such a way so as to be able to demonstrate it works. I have to detail what buttons to press, what to type, what you see on screen, all in huge verbose, repetitive documents. It literally has so little to do with programming that the business grads who also recently joined are helping out.
Now, if this was for two weeks or something, I could live with it. Who better to do this job than the inexperienced graduates? But, it's not two weeks. It's three months.
Feeling this way, part of me feels very selfish and spoilt - hard work isn't easy, someone has to do the boring jobs, and it might as well be the graduates because they're cheap.
But the other part of me feels slighted, and lied to – I turned down some very promising opportunities for this job and now we won't even start TRAINING until January.
Honestly the work is so dull I think eating the keyboard would be much more fun at this point. I've lost all motivation and excitement for the job, and I'm finding it very hard to remain upbeat to my superiors.
So, what do I do? Do I suck it up and wait it out until the “real” programming starts? Or do I approach my manager and ask if I can help the “real” developers? If I ask my manager, how do I approach it so that I don't sound like some entitled graduate who thinks he knows everything? I understand that if they put me on the dev team, I'm going to be the least experienced developer in all liklihood, but I can learn pretty quickly and out of the other graduates I've probably got the most experience. I want to come across as eager but not entitled and self-absorbed. I feel a mix of emotions: I'm pretty upset as I feel I was lied to, but I also feel guilty for feeling this way because I know jobs are hard to come by. But I don't know if I can deal with it for 3 whole months. I was so excited now I'm just :(
[QUOTE=Map in a box;48861161]As much as I love Lua, I'd rather make something from scratch than be associated with King.. and they're not using LuaJIT :([/QUOTE]
While I agree; King's milking techniques are horrible, they're still a very successful company with 1500+ employees and getting hired by them is pretty difficult. The developers work in small teams (10-15 man) and they can pretty much choose a developing style that they enjoy. They don't have any pressed deadlines so people aren't forced to work overtime, etc.
Sorry, you need to Log In to post a reply to this thread.
