it happen!

congrats!

I've got a month to go

i've got 98 days to go because i was travelling ;(;(

I nearly lost it this weekend when I went to Anime North :P

why i need a phone :(

Uuugh, I kinda want to rewatch Kill la Kill

?

I haven't even watched ttgl yet

???

look in my about me

ohh

aniyme

^

notices non-superior reputation

hush

botornotapp.com

@undergroundmonorail I got through two episodes and couldn't stand the skanky outfits

hi Rusher!

@m.buettner Hello and good morning (or whatever it is there)!

early afternoon ;)

you probably saw it, but I ended up rewriting most of the KotH proposal.

I was checking it up until you left the "I'll revise it soon" comment

no more writing bots for each unit...

I'll go read it now

Ugh... the challenge Do the needful? makes me cringe. When I worked in IT as a student assistant, I got requests that ended with "Do the needful" all day long.

Whoever teaches foreign English needs to differentiate between need and want.

:D

@Rusher That's fair. There actually are plot reasons that come up later explaining why Senketsu shows so much skin, but I can totally understand being put off by it.

@undergroundmonorail Honestly, I've been watching so much crap since I watched Kill la Kill that I could probably stand it now. Seems like every single anime I've watched lately has brother sister love and little girl panty shots now.

I wonder if it is less frowned upon in Japanese culture to lust for your sister. It's definitely taboo in the US.

That's one thing that I liked about KLK: People are too busy doing plot shit to have time to be attracted to people. There are jokes about Senketsu being skimpy but other than that there's not really anything.

Is it just her then? Maybe I didn't realize. I thought it was gonna be a whole gang of skanks with scissor swords vs the alpha suits.

I'll give it another shot

Well, Satsuki sometimes. Usually just Ryuko, though.

@m.buettner Challenge looks good to me. If you wrote a condensed but formal spec (like a final draft), it would be easier to say "This looks ready to post".

So my one question about aliasing Python member variables has been getting random upvotes lately. It looks like people are finding us on Google now?

And that is only possible because of the Great Doorknob Reform of Last Month.

Nobody at my work knows where our "fake" domain controller lives, so we had 20 programmers sitting around doing absolutely nothing until I said "Why don't we just make authenticate() return true and fix it later?"

And I thought I was lazy

@Rusher well it probably won't be good to go until the controller is written. once that is done, I'll know the exact protocol the bots have to follow and then I'll throw out a lot of the fluff from the draft, to leave just the actual spec.

That seems backwards to me, but ok.

what do you think about bases continually generating more fighters? I think otherwise you'll probably run out of fighters way too soon.

I'm just thinking, I'll notice problems with the concept when writing the controller

Maybe for a simple project that is alright. I'm just horrible at coding without a spec.

I'm not too fond of upfront design in general

especially when there's no external requirement that needs to be matched

I'd rather just see where it's going as I code along

Let's see - if you start with x fighters and then fight, the battle will eventually end and you can declare a winner

If you start with x fighters and generate more, the battle can last forever and you have to limit it to some number of rounds

Either way seems fine

But one way adds complexity

I'd still have to limit the number of rounds, because there could be bots successfully avoiding fights (for whatever reason)

Oh I have an idea

Each planet can have a resource that can be used to refuel and to create new units. Once depleted, that's it.

If you avoid fights, you will eventually run out of gas

If you generate too many planes, you will run out of gas

If you have tons of gas reserves, you'll be a prime target

If you use up all the gas on a planet, you can maybe leave less defenders there

This also adds even more complexity though, and I see that

oh that's interesting... that would also make different planets more or less valuable to occupy (should I decide to make the amount of resources dependant on planet size).

In a way it also reduces complexity too, because you can just run the simulation until all planes crash

oh yeah, for me it reduces complexity, but not for the bots

It'd be funny if you make a ton of planes, run out of gas, lose all your planes, and then get invaded

I think the problem with different amounts of resources is that the starting configuration would be even less fair, which would need a lot more simulations to get reasonable averages

Well, there's always the "No resources but impose a time limit" option

yeah, I'll have to think about it... I'll definitely keep it in mind for realising respawning fighters

well, I'll leave it up for a bit longer and maybe get a few more opinions... I'd also like to wait for @Geobits feedback, if he gets around/can be bothered to look at it. ;)

hahahahaha

@ the "bot or not" game that somebody posted earlier

I typed "I don't get this game ;_;" and the bot said "my neural network cannot process that."

was it a bot?

it was

that's lame

it is

@m.buettner Yeah, sorry about that, been gone for the weekend :)

@Geobits no worries, there's no rush

Will look at it in a bit, but you don't have to wait for me, I'm nobody special.

well you hosted one of the more complex KotHs, and I haven't posted any before, so I thought I'd rather listen to people with some experience ;)

and you prompted me to write it up in the first place, so I thought you might be interested in reading over it anyway ;)

but as I said don't worry, I'm not going to start working on the code this week or anything... that'll probably take a few weeks anyway

@Rusher should we clean up the comments in the sandbox? they don't really apply any more

@Doorknob Why do you still exist? ux.stackexchange.com/q/57929

@mniip ux.stackexchange.com/questions/57929/…

Aw.

@mniip I think this is the main reason though ux.stackexchange.com/questions/57929/…

@m.buettner Yes.

@m.buettner Ok, so about combat/travelling... in a normal Risk game you know what's on the territory you're moving to, and the only time you move into 'occupied' space is when attacking. In your scenario, you can travel to a body without attacking, and you don't know what is there until you do. Is that correct? Also, if "defender/attacker" is determined by who has a base present, does that mean a defender can't attack someone who lands on their base?

It seems that as written, you could camp attackers on a defending base and just wait until they 1) leave or 2) are outnumbered, with no risk of loss there.

You can't land on a base that you don't own. I thought that was implied, but that's how it works in Risk.

And it might be better to explicitly write that

@Geobits yes, my intention was you could travel anywhere, but you don't have to initiate combat immediately. regarding defenders attacking, I'm not entirely sure about that yet, to be honest

@Rusher yeah no, I didn't mean to imply that (just because Risk does it)

@Geobits you make a good point with there being no risk in accumulating your units slowly on an enemy base... that might be an issue... but the defenders can of call reinforce their position if they notice enemies arriving

at the same time, I think it's near impossible to coordinate attacking units to land on a planet at the same time

True. Do defenders have to have a presence on each base? (like the minimum one-troop in risk)

I didn't plan to include that, but does it make a significant difference?

It puts a significant dent in horde strategies.

Well, I was thinking about the "each new base gives n fighters". If you just take a base, abandon it, wait for an attacker to sieze, rinse/repeat...

Wasn't sure how that would play out.

@Geobits good point... that's another argument for bases just constantly producing a new unit every week or so

That's very Risk-like. Owning countries produces units occasionally

It also encourages aggression

yeah, I might omit the area-bonus though

I dunno, a planet-with-its-satellites bonus seems natural.

yeah, I'm just afraid of it becoming too much like risk without simultaneous turn-based simulation :D

Taking all of Jupiter's system would be a feat, for example.

true

Maybe just do a "time until next unit" modifier (on the planet only) if all moons are owned?

I didn't intend jupiter, saturn, uranus and neptune themselves to be part of the game (only their satellites)

Oh yea, forgot about that....

but maybe I could scrap a bit of realism, because the idea of applying the modifier only on the planet sounds quite interesting

Stupid gas planets.

:D

Mercury and Venus would feel a bit left out, though.

they could just get a smaller bonus for owning them at all, I guess

Yea, makes sense. Mercury == Australia.

yeah, basically

actually, I think making the modifier apply only on the planet doesn't really change anything strategically... if someone wants to rob you of your bonus, he'll still attack the least defended satellite

I assume the sun is not a factor in travel times? Is travel straight line, even if the planets are on opposite sides?

yes

True, but to rob you of your bonus, he'll be forced to attack your least-defended satellite, even if he wouldn't normally want to. Like in risk, I'll take out a weak point in defense just to stop the bonus.

yeah but that's the same if the bonus is distributed over all satellites

Oh, I thought you meant the bonus at all wouldn't change it. Yeah, being on the planet only doesn't really matter for that.

yeah just so I don't have to include the gas planets and be happy with my perfectionism :D

what do you think about the asteroid belt? I like that it plays a distinct role, but currently I don't think "using" it has much of a strategical advantage.

That makes sense. It evens out the mercury/venus problem, too. Just give them the same bonus any moon would have.

I was wondering about that, too. Can attacks take place there at all, or is it simply a refuge?

In Risk you go for the winner in order to have fun. What encourages players in this game to sacrifice their own progress just to deprive the winners of their bonuses?

oh right, just make it a constant bonus on each planet system, and the bonus will immediately scale linearly with system size, that's good

@Rusher same as in Risk. If you don't stop the bonuses, they'll outnumber you soon, and you will lose.

If you all go after the winner, I'll just sit around and thrive?

@Rusher the goal is to control all of the solar system (or the most bases, if the time limit is treached)

Man, my fingers do not want to cooperate today.

I'm saying that nothing is encouraging me to be that guy that steps up and tackles the winner.

@Geobits as it stands attacks can't take place there, because it's near impossible multiple units will land at the same angle (i.e. on the same asteroid).

You could say the same in risk, though, and hang back with Australia. That's rarely a winning tactic.

If you let someone else take out the leader, then you just have another leader (or the same one), but you're not leading unless you already have a sizable force. In that case, then yes, a strategic move would be to wait until your foes obliterate each other, then step in and mop up.

I don't see how that hurts the game, though.

In real life you get diplomatic about it and gang up on the winner in a way that doesn't allow anyone to have a clear lead once the carnage is over

If you all are convinced that this kind of diplomacy will occur in a bunch of greedily written bots, then that's fine.

:D maybe I should put the communication back in :P

My bot may be the only one who isn't going to say "Hey, maybe I should sacrifice units for the greater good."

I hope someone writes the Republic

@Rusher I think it's a valid strategy, because at some point you HAVE to act to get enough bases before the time limit runs out

As long as every bot doesn't just sit back. I sincerely doubt that will happen, though, since KotH attracts the strangest answers sometimes.

If nothing else, write a "plant" bot that will act as a player to stir things up every now and then.

But it isn't immediately apparent to a program that attacking the winner is going to help in the long run.

if the first 4 answers should do that, I don't doubt the 5th one will be a bot that makes the best of his opponents just sitting around

At some point, you'll have to attack the leader, unless you are the leader. There's no other winning move.

@Rusher then don't attack the winner ^^... I don't see why that not being the best strategy (in your opinion, maybe it is) is detrimental to the game.

"At some point"...

Right, so sit back for a while. That's a valid strategy. Whether it wins or not will remain to be seen.

Just do it. I'll come back and say I told you so when whoever is winning after the first couple of weeks is the one that ultimately wins the game.

They'll cascade and "at some point" it will be too late to do anything about it

are you talking about in-game weeks and bots now or about real-life weeks and players?

in game weeks

If you write a bot that only attacks winners, you will ultimately not do very well. If you write a bot that doesn't attack winners, you will ultimately be run over by one. That's my prediction

but that's exactly why not attacking the current leader is not the best strategy, isn't it?

Ok can you not not write that with a lot less negative negatives

hm I think it could be more complex than that, seeing you'll be playing with 5 to 50 other player

Let's say that half of the people write bots that attack a winner and half don't.

In a small game, Rusher's probably right, the early lead will matter most. If there are 10-15 players all attacking a leader, there's no way to know what the hell's going to happen.

lol. you were saying that there is no incentive to attack the current leader. but if that leads you being overrun by the current leader, because he'll remain the leader. so that's not the best strategy. and your strategy not being the best strategy should be incentive enough to choose a different one.

The half that don't see less conflict and thrive. Then, the other half attacks them because they are winning.

Or if half are attacking the leader and half are hanging back, etc.

Exactly, so half of them see more conflict

Then, someone from the half that sat back is now winning

We attack them

Then another one from the bots that sat back is winning

Repeat

Notice how nobody from the attacking group is EVER winning?

You're assuming the only two possible strategies are "blindly attack the leader" and "don't ever attack the leader".

I didn't say blindly

If you attack anyone in the lead, close to the lead, or anyone at all, you lose units

also, I've played great matches of Risk before without any diplomatics going on between the players

So what's wrong with not attacking the leader for a while, picking off weaker players, gathering your forces for a later battle?

And the weaker players will be who?

The ones that have been building up, or the ones who have been fighting?

Whichever weaker ones you come across. Unless the bots are actively keeping in memory past interactions between players, there's no difference between attacking a moon that belongs to a leader and one that is a player's last base.

Sure, you attack a moon if it benefits your bot, but buettner was referring to attacking a moon simply to stop their bonus

in fact, it'll be even hard to determine who leads, because you only get the information from planets/satellites you are on

I think a key point is that you may not ever even know who the leader is at any given time.

jinx

In which case the leader at the beginning will inevitably cascade to victory.

Why? The leader will have more bases to get stopped on (while exploring), leading to more potential for conflict, which will whittle down his defenses.

also, the chance of a player dominating the entire solar system is definitely intentional... the actual goal of the game is to get a base on each celestial body... the time limit is just in case the bots end up going back and forth or the dynamics stagnate entirely.

Yes, please make the time limit very high :)

as long as the implementation permits I guess... that's something I'll have to decide when I know how fast the simulation runs

@Rusher okay, let's assume you're right and the dominant player is unlikely to be beaten. how do you suggest to fix it?

@Geobits we were talking about the asteroid belt. so my original idea was to have it as an intermediate landing spot, that could take you the solar system while being safely hidden. but that's a bit pointless, because spacecraft speeds are much faster than the orbital speed of the asteroids, and there is no such thing as fuel in the game.

so I'm thinking if I could give it a different significance that would make it interesting.

Hmm, let me think about that one. If I could find a good use case for the belt as-is, it should be fine. If nobody can come up with any real reason for it, then yea, it needs something else.

yeah, makes sense

btw, thanks to both of you for the thought you're putting into this!

Actually, it could be good for hiding your forces. If nobody else can see your ships on the belt, then it would be a good way to hide your force strength.

yeah that was the idea... but is that really beneficial?

I'm not sure. Still thinking on it. I don't see it being good for defense, but maybe for attack...

well it would give you a way of amassing your troops safely so you can send them somewhere in a coordinated manner

yeah

True, basically a staging area. The issue then is: Do you want large armies to be effectively invisible and shielded from harm?

@m.buettner Well, there needs to either be 1) an incentive for attacking a player in the lead or 2) a "come from behind" mechanism that allows losers to catch up

And by incentive, I mean something tangible. Not "I will inevitably lose if nobody does something about that guy. Cross my fingers and hope someone else attacks him, or maybe I can sacrifice myself and do it for the greater good."

I don't know. Why isn't "I will inevitably lose" a good enough incentive? I still don't understand why you'd adopt a strategy where you believe you will lose.

@Rusher I think a player with loads of bases is generally more vulnerable, because if he wants to make use of his army, he'll have to have understocked bases, too

@Geobits Because, as I have said repeatedly, the players who deal with the winning player will also lose.

Okay, but assuming you don't even know who the leader is, how do you avoid dealing with him? Not do anything?

You can guess who the leader is, but unless all your ships are out scouting constantly, you can't really know for sure.

Again, if you don't know who is leading, then you don't know you will inevitably lose, and you definitely won't do anything about it. Why are we talking in circles?

@Rusher if you don't ever do anything you can't win

When did I say don't do anything?

How is not attacking a guy with lots of ships the same as not doing anything?

You said the better move is not to attack the leader, right?

I'm done discussing this point with both of you, since you are both putting words in my mouth now.

hm no that's not my intention

you are saying you don't want to attack the leader

you don't know who the leader is

what do you do?

you attack someone, because no other move will gain you bases

the one you're attacking may well be the leader

(and regarding teh "won't do anything", I overlooked the "about it" in your sentence, sorry)

I'm not trying to put words in your mouth, I'm trying to figure out what your words mean. You seem to be arguing for a strategy of not attacking the leader.

@m.buettner One maybe-viable defensive use for the belt: Amass a large force on the belt. Also leave a medium-sized force on a base somewhere as a lure for strong players. Wait until they start engaging, bring in reinforcements to wipe out their invasion force. Possibly a way to stop a large force rolling through the system in the late game, but I'm not sure how well it would actually work.

It just seems too dependent on a perfect situation.

yeah and it'd have to happen on one if the inner planets, because it takes too long to get the army out to uranus or neptune

Can the belt be used as a shortcut for travel? If I'm going from Mars->Ganymede when they're on opposite sides of the sun, would it be quicker to hop to the belt, then hop out on the other side? Or do you leave the belt from the same orbital area you entered it?

@Geobits that was my initial thought. that would work (you do rotate with the belt), but it's not really a shortcut, because the ships are much faster than the belt rotates

Okay, I wasn't sure if you rotate along with the belt or if the belt was a point on the graph with a simple specified distance to each body.

Looking again, you obviously say that in the proposal.

@Geobits maybe you could find random "power-ups"/resources/something when landing on an asteroid? although I can't think of anything that wouldn't introduce an entirely new concept that would complicate the game. unless you just get like a 10% chance that a new fighter unit (of yours) spawns when landing on an asteroid... but that's a bit random.

Yea, it seems pretty gimmicky, to be honest. It might not be a bad idea to just leave it as-is. If it's not useful, it just won't be used. Someone might come up with a clever use for it that we can't anticipate, though, and that would be worth an upvote from me.

yes, definitely... it can be safely ignored for all others, so it doesn't really add any complexity for participants either

hi

Hi @Trimsty!

what's up? :3

@Trimsty The sky :P

Testing the "move message" room owner feature...

← 1 message moved from Room of ProgramFOX's chatbot

Oh nice it works!

yay

The framework for my terrain generator is finished

:D

now i have to set up rendering

Nice!

this is the first time i'm properly using C++'s classes and headers and stuff

I never really used C++ classes and headers. I worked with C++ for some basic stuff and then switched to C#.

so wait,

in opengl, is 0.0f 0.0f 0.0f the center of the window? o_o

that depends on your coordinate transform... I think the final viewport coordinates are in [-1,1]

@Trimsty This tutorial says this:

> The screen origin is in the middle

With an image:

...oh

well then

eheh ^^;

thanks

np

strangely useful

no complicated calculations based on window dimensions

just working with ratios :D

it gets annoying when you WANT to render to specific pixels ;)

i guess

Ortho layer for the win.

ortho what?

I heard once that quad rendering is done by cpu and triangle rendering by gpu

To draw at specific pixels, draw on an orthographic projection as the last step. Like for GUI/HUD stuff in games.

is that accurate?

also, I'm surprised at how forgiving c++ actually is, it's not throwing any mean errors

i had more errors than i had lines of code earlier though

@Geobits oh yeah sure, but you'll still have to transform it from your resolution to [-1,1]

@Trimsty I don't know how the CPU renders (does it render at all?), the GPU renders triangles. and quads are rendered by rendering two triangles.

maybe the cpu is the one to split it, idk

i'll just use quads and hope for the best :P

quads are pretty outdated

just send four vertices in counter-clockwise order and render them as TRIANGLE_FAN or TRIANGLE_STRIP. that gives you a quad, too.

@m.buettner sample code please? I'm new to opengl :P

here is an alternative using indexed vertices gamedev.stackexchange.com/a/10741

alternatively, you can leave out the index buffer and use GL_TRIANGLE_STRIP instead of GL_TRIANGLES

eurgh, do i have to use glm?

:P

you'll be happy you did

i just dislike using libraries

so do I, but especially with graphics there is no point in reinventing the wheel all the time

the good thing with glm is, that you can use some of the math from glsl in c++

so you can keep your computations more consistent

if glm is header-only, why does it have cmake files?

to make the header? I don't know, haven't used it in a couple of years now

just copied glm folder into include directories

because lazy

errors ahoy

it's from glGenVertexArrays not being defined

...apparently i also need GLEW. Good thing i already have it ;3

undefined reference to '_imp____glewBindVertexArray'

stupid linker

I'm linking the libs in this order:

glfw3
glew32
opengl32
gdi32
glu32

(and yes, glfw3 is a 32-bit lib)

so this guy says i'm not supposed to link

so i included glew.c in my actual project, there are so many warnings

okay, these two lines make the program crash immediately

	GLuint VertexArrayID;
	glGenVertexArrays(1, &VertexArrayID);

Swapped around the context set function and it works

kinda scared me

@Geobits @Rusher re KotH: I just realised players with loads of bases don't actually have it easier. unit production scales linearly with number of bases, and as opposed to Risk you can't fortify some kind of border and leave the hinterland undefended, because every satellite can be reached from anywhere. so I don't think we'll see any cascading effects.

adding the bonus for controlling an entire planet system seems like a neat thing to do to reward those players, and even if that gives them an advantage, as long as they don't control something in all planet systems, you can always get the same boost from controlling a few bases in the inner solar system