Game Development

GnomeSlice

13:01

lol

Jon

13:39

Anyone know how to "Square out" a pathfinder?

GnomeSlice

@Jon I know how to cut grilled cheese into squares

Jon

Alex M.

square out?

Jon

like for example.. my a* implementation is too good... i'd rather have it focus more on going one direction, then the other

instead of switching back and forth from right, up like above.

Alex M.

so, left-left-left-left-left-up-up-up-up-up?

Jon

13:45

essentially.

Alex M.

not sure how this would fit in your case but:

Bresenham's line algorithm

The Bresenham line algorithm is an algorithm which determines which points in an n-dimensional raster should be plotted in order to form a close approximation to a straight line between two given points. It is commonly used to draw lines on a computer screen, as it uses only integer addition, subtraction and bit shifting, all of which are very cheap operations in standard computer architectures. It is one of the earliest algorithms developed in the field of computer graphics. A minor extension to the original algorithm also deals with drawing circles. While algorithms such as Wu's algori...

Jon

naw..

I would need some form of pathfinder, that prefers straight routes.

Alex M.

since you're A*-ing I guess you have obstacles

if I were to try to do what you wanted, for a first draft I would:

Lasse

blogs.msdn.com/b/dotnet/archive/2014/04/07/…

Alex M.

1) Enclose found path in a rectangle.

2) Try to go to the right/left until reached the corner of the rectangle or an obstacle.

3) Avoid the obstacle.

4) Repeat.

Jon

13:51

Yeah...

A* is probably not the solution here

Kevin van der Velden

There's also non A* algorithms that avoid it

MickLH

i think A* will work fine

Alex M.

@Jon why not?

MickLH

add some kind of heurestic that makes going in the same direction cheaper than changing direction

maybe lol

Jon

Well.. Yeah.. kinda defeats the purpose of A* lol

Alex M.

13:53

what Mick sorta makes more sense

Jon

But yeah, I would have to create a separate implementation of A* just for this

Alex M.

well tbh you're not really asking for the shortest path anymore

so you defeat the purpose of most pathfinding algorithms in their standard form

Kevin van der Velden

This gamasutra article has some things that can make A* generate more realistic paths

Alex M.

@KevinvanderVelden that's what he doesn't need

Kevin van der Velden

OOh! I found the pathfinder algorithm I was gonna recommend: aigamedev.com/open/tutorials/theta-star-any-angle-paths

@AlexM. hmm true bleh

Misread

I thought he wanted to make it smoother

Jon

14:01

ok... so maybe I don't need to use pathfinding here

its a start i guess

Kevin van der Velden

Wait, you're going from the center of each black thing to another black thing only?

Then just create a simple graph and A* it

Jon

at the moment yeah

i just dropped the A*

Kevin van der Velden

Use Manhattan distance heuristics rather then euclidean

Jon

I'm not seeing your vision

Kevin van der Velden

@Jon well jeah, A* over the 2D map pixel for pixel, I'm saying create a graph with vertices for each black point and edges to each other black point

GnomeSlice

14:13

@Jon indiedb.com/members/open-sketchbook/images/…

you guys are so smart

Jon

@KevinvanderVelden I already did this

The issue is A* is too direct

Ends up with pathing that is too unpredictable

Kevin van der Velden

From the looks of it you did it over the pixel/tile map, correct?

Jon

No..its a dungeon data map

the pixels are just drawn after the fact for debugging

Kevin van der Velden

Then I don't understand what you're doing currently

Jon

lol.. well if I run A*, I get this

Alex M.

14:17

@KevinvanderVelden instead of a straight line towards the target he wants detours

Kevin van der Velden

@Jon and what I'm saying is you don't "just" run A*, A* is an algorithm that works on a graph, how are you constructing the graph?

Jon

Right. Instead of A* pathing, I want a more straight instead of direct path

user92578

Hey there!

Kevin van der Velden

Hey

Or if you don't know what the graph is then I already know the answer

Jon

Lasse

14:19

path = block1pos - block2pos;
for(int x = 0; x < path.x; x++) { // horizontal line }
for(int y = 0; y < path.y; y++) { // vertical line }

like that maybe?

Jon

Dropped a*, but now the pathing doesn't care if there is a room in the way... which I guess is fine.

I'll go with this for now

and most likely refine it further as I go along

Alex M.

you'll end up writing your own pathfinding algorithm that way :D

Lasse

and then on each dot test if it's inside another block, and if it is then take some other path (vertical line before horizontal)

Almo

@ShotgunNinja 23-2-11 cait, team got 47 kills last night. :D

user92578

@Pip I'd like to test Malware Tycoon!

Kevin van der Velden

14:25

I'd like to test my Malware!

Oh wait, we were talking about something different

user92578

lol

user92578

I'd actually really like to try hacking or something :)

Lasse

oo evil hacking

user92578

or just write a bot that spams my friend's emails :DDDDD

Lasse

why use bot, when you can use botnets

it's more efficient

user92578

14:28

But, my codes can't use networking

user92578

at all

user92578

so nothing works

user92578

firewall blocks it

user92578

and I don't have the password to allow it

user92578

maybe I need to hack the firewall first

Lasse

14:28

yes, the software ones are quite easy

user92578

really? well I've heard that the Window's firewall isn't good at all...

Lasse

well just bombard it with packets until it stops responding and windows kills it

the downside is that windows most likely will kill your app bombarding the firewall first

user92578

the bad thing is that I'll probably break something...

MickLH

the solution to that is figuring out what causes the firewall to "think" really hard

hand craft a dirty packet that will choke it as hard as possible, then spam it with that, standard DoS attack

user92578

and bam I've killed my firewall and now all the other malware can infect my computer

user92578

14:32

lol

Lasse

oh yes

but you have unrestricted access to the network!

Kevin van der Velden

@Tyyppi_77 do you have a router? Then you're still 99% safe

user92578

not sure at all

Lasse

That's why people who know about shit uses linux

@Tyyppi_77 what's your ip address?

user92578

localhost

Lasse

14:34

open cmd and type in ipconfig and you should see it

user92578

it's localhost

Kevin van der Velden

@Tyyppi_77 -____-

Lasse

localhost isn't ip address, it's hostname

if the ip address starts with 192.168 or 10.0 or... uhh.. what was the third one? then you wil be behind a NAT box

or router

user92578

it starts with 192.168

Lasse

then you are most likely safe without the software firewall, which is blocking only traffic going outside

Shotgun Ninja

14:38

@Almo Nice!

Almo

:)

lulu support is pretty awesome

also learning each shot to shoot at the target most likely to die

that means switching target as the enemy adc/support shift position

just keeping target on the adc often means the support sets up to kill you

Shotgun Ninja

well yeah

establish a line.

anyone who crosses that line gets fucked with.

Almo

:D

what you think of GA as last item

I checked a pro build, and they seem to use it

Alex M.

(/ .□.)\ ︵╰(゜Д゜)╯︵ /(.□. )

fyeah

user92578

to be honest, I don't know anything about networking

user92578

14:47

maybe I should learn some

user92578

code a network simulation since I can't actually network

Alex M.

list comprehensions in python are pretty neat

user92578

yeah

Alex M.

self.tiles = []
for i in range(n):
	self.tiles[i] = []
	for j in range(m):
		self.tiles[i][j] = 0

user92578

they are some awesome

Alex M.

14:48

becomes self.tiles = [ [0 for x in range(m)] for i in range(n) ]

Lasse

thats..uhh.. unreadable

Alex M.

not really

Lasse

yeah well might be because I don't know python

Alex M.

it's really just mathematical notation

v = [ x for x in range(3) ]

gets you [0, 1, 2]

it's equivalent to V = { X | X in [0, 3) }

reddit.com/user/InconsolableCellist

Almo

need a free flowchart tool? yworks.com/en/products_yed_about.html

really works well

Jon

14:54

not so damn bad for an hours worth of work if i do say so myself

GnomeSlice

@Jon :o that looks sweet!

Almo

:D

only complaint: the double corridors. fucking great for one hour.

Jon

Custom pathfinder a la jon

I kinda like the double corridors, ehehe

Almo

ah! even more win then

:)

Jon

super fast generator if anyone wants it, lol

GnomeSlice

14:56

sparkrising.com

Jon

maybe i'll make a blog post about this dungeon generator, since everyone else has one

Almo

:D

GnomeSlice

@Jon Make a gif showing how it works

like that

Jon

can't see that image

GnomeSlice

blocked?

Jon

15:00

yeah hold on i'll use the other network

oh

yeah

that gif is awesome

GnomeSlice

There, that should work on both now

MickLH

I fucking relied on my text editor to keep 2 version of my code in memory

Everything was going smooth, until I ran the god damn buffer out

Now I wrote what I thought was the same code, but it's got the original error that I fixed in the undo buffer ;(

Almo

sounds like you need to write a new text editor that won't fuck that up.

Evan

:D

Jon

That is a very intelligent level generator

I need spacing between rooms, so it wouldn't work for me

GnomeSlice

15:05

Jon

my algorithm is pretty simple and dumb

1. Define an area.
2. Generate random rectangles inside area.
3. Connect Rectangles

GnomeSlice

Seems to work!

Jon

Rectangles are checked for collisions, and discarded if hit

When generating rectangles, I have

if (count > RoomCount * 3)
                    break;

GnomeSlice

blank stare

Jon

Yeah.. I don't know how to stop it if there is too many rooms

and it goes in infinite loop

Icy Defiance

15:09

eeeeewww break statements D:

GnomeSlice

cap the room count to like 20 or something?

Jon

Yeah, I have a room cap. works fine

ie. RoomCount

if I loop over 3X the roomcount, I stop trying

Almo

what's wrong with break

Jon

it breaks out of the for loop

I use breaks when they are useful.

and save time

Icy Defiance

@Almo just that approaching the problem a little differently to avoid using break will (almost) always result in cleaner code and easier to follow logic.

Jon

15:11

I think the issue with people saying break; is bad, is because they were told it is bad

Almo

^

Jon

I posted this code because I don't think it is optimal.

Almo

though gnome is not necessarily doing that

Icy Defiance

it's bad in the same way goto statements are bad. it's just the degree is less.

GnomeSlice

@Almo I just linked to a gif.

I have no idea what they did.

Almo

15:13

I meant that you weren't "saying break; is bad, is because they were told it is bad"

MickLH

I think any coding construct can be bad

Almo

necessarily

Icy Defiance

that was me, not gnome

anyway I didn't mean to derail the conversation. it was just a side comment.

MickLH

NO! I must defend the honor of the break verb!

Icy Defiance

D:

MickLH

15:15

I'm joking but I do think it actually really helps readability in some situations

Kevin van der Velden

I for one am a great fan of break statements... especially in switch statements

Jon

ok

i was looking for an example, and found one

Icy Defiance

well yeah switch statements kind of require them, lol

Evan

Not necessarily

fall through switches is seen frequently

Icy Defiance

usually

Jon

15:16

bool validated = true;
foreach (Room room in Rooms)
{
	if (room.CollisionRectangle.Intersects(r))
	{
		validated = false;
		break;
	}
}

Evan

they are questionably reliable though

Jon

what would you do with this, code review time.

MickLH

what is r ?

Evan

Indeed

Jon

a rectangle

Kevin van der Velden

15:17

@IcyDefiance that's kinda the joke =p

Evan

Also, validated is somewhat vague, though a lesser issue

MickLH

It seems too clear and simple to be wrong :P

GnomeSlice

Shattered Planet's ready for your Android explorations! Get downloading! https://play.google.com/store/apps/details?id=com.kitfox.shatteredplanet http://t.co/STKHX85VwH

Tweeted by KitfoxGames on April 10, 2014 at 2:29 PM

woo

Jon

I think it is a fine use of a break;

MickLH

but with that said, some kind of BVH would probably be easy as hell to implement

Evan

15:19

if you REALLY wanted to, you could restructure it to eliminate a conditional check. by using a regular for loop that checks validated as its conditional, and the loop itself would use the results of the intersection test to set validated

Thats just unnecessary though

and less obvious

So you should probably not do that

Alex M.

@Jon it is

Icy Defiance

for (int i = 0; i < Rooms.Count && validated; i++)
    if (Rooms[i].CollisionRectangle.Intersects(r))
        validated = false;

Alex M.

don't do that

user4704

That's basically a break, but in a less common style.

Icy Defiance

there

Jon

15:20

that would also work

Evan

Dont even need to do the if check for the intersection, could just assign validated to the inverse of the intersection test

Icy Defiance

@JoshPetrie yeah you're right. I'm not entirely serious about that solution, though I have used it before when I felt lazy.

Evan

:D

user4704

Rewriting that as a while loop would read in a more expected fashion.

Icy Defiance

good idea

Jon

15:21

interesting

Alex M.

@IcyDefiance at least add the brackets, as jon seems to already do

no offense, but removing brackets to make the LOC smaller is what I call the idiot's refactor

Icy Defiance

nah in that case I'd remove them because I think it looks better.

int i = 0;
while (i < Rooms.Count && validated)
{
    if (Rooms[i].CollisionRectangle.Intersects(r))
    {
        validated = false;
    }

    i++;
}

Alex M.

I hope you at least have the decency to add them if the previous programmer added them in the module you're working with

Icy Defiance

oh yeah. consistency is good.

Jon

looks good to me

Icy Defiance

15:24

actually I still like the for loop better than the while loop here, because the scope of i is limited.

Jon

Yeah, I prefer the for loop also

easier to read in my opinion

Alex M.

I think breaks are fine in a lot of cases, and improve both the code's layout and its ease of understanding

Jon

Also, if you are using VBA for certain things, especially in Excel, breaks are pretty crucial (or exits)

MickLH

I still think the foreach was the cleanest

Alex M.

I always do this

public void Update()
{
	if (condition)
	{
		return;
	}
	rest of the update logic;
}

when update needs to be stopped on a particular condition

Kevin van der Velden

15:27

@Jon but then, you shouldn't be using VBA for anything so that really isn't an argument either way

Jon

I do a lot of if (condition) return; in functions

not sure if that is bad, but I love doing it

Alex M.

it's readable to me

user4704

It's fine.

MickLH

Lol oh come on it's not bad, there needs to be a reason for it to be bad

Alex M.

and it's readable to any programmer who wrote more than 20 lines of code

user4704

15:28

Very old-school programmers will frown upon having multiple-exits.

Alex M.

that embed-valid-condition-in-for and whatever

user4704

But that's largely a perspective developed from extensive years of C.

Alex M.

is just academic bullshit

that is pointless

MickLH

Very old-school programmers are outdated and irrelevant like the betamax tape compared to the DVD

Still complaining about VHS

Alex M.

the first version was perfectly understandable

I don't even understand why you brought it up

Evan

15:29

Old-school programmers bring a perspective to development that new programmers lack

2

GnomeSlice

Today's Listening

Alex M.

@Evan but since new programmers are just fine

Evan

I have found some of my favorite programming books are "old and busted"

Alex M.

it must mean the perspective isn't really all that useful?

MickLH

@Evan Not always, I've actually run into the problem of the majority of them believing this is the case but then not actually delivering on the fresh perspective

Icy Defiance

15:30

yeah that I do as well, even though I'm told many times that multiple-exits is bad, because the alternative is often something like
if (!condition && !condition && !condition && !condition && !condition && !condition)
{
//logic
}

return;

Evan

Okay, well I guess your mileage may vary :P

MickLH

I've never not had an old-school programmer complain about efficiency to me

Evan

My results with old-school programmers has been good.

Icy Defiance

bah damn formatting

Alex M.

I bet an oldschool programmer with his perspective would do wonders with HTML5

Jon

15:31

LOL, oh well... some older programmers here tell the juniors to not use newer libraries, because they don't understand them.

linq, regular expressions

Evan

There are exceptions to everything

user4704

That just means they are bad programmers.

MickLH

BUT! Despite each one complaining about efficiency, my efficiency has never been below that of their on the same problem

Jon

I don't work with them, so I don't have to deal with that :)

user4704

Programmers from an older generation are not good programmers by virtue of their age, after all.

MickLH

15:32

I've had these old guys so sure that they are better, that they stir up management to do a performance shootout between our implementations

Icy Defiance

@Jon both of those are incredibly useful, but also incredibly overused. like here: stackoverflow.com/questions/1732348/…

MickLH

I've never, ever lost one of these, and statistically, the more sure someone is that their code is better than mine, the slower their code runs relatively

Also these old guys hated HTML5 because "Javascript is too slow and bullshit"

I guess somehow it's more efficient to have the server sort a page, and then re-load the entire page

Sorry for the rant, I just write off "old school" as retarded these days, because every time I learn the latest thing with coding, it's better than the last thing

Alex M.

@JoshPetrie yay, someone got it

Kevin van der Velden

@IcyDefiance aah good, I was gonna look for the ZALGO reply but that was this question :)

Alex M.

the whole discussion is aimed at the wrong target

programming is about the technology

and each substantially different technology requires a substantially different perspective

if you are not capable of grasping that perspective, you're just a bad programmer

regardless of your age

and should switch to a different tech

BUT

Evan

15:36

Well, there certainly is an interesting collection of perspectives here.

Alex M.

if I were to choose between an oldie and a newbie if it were about learning a new technology

I'd choose the newbie because he'd be less likely to try to do FORTRAN in JavaScript

MickLH

I appreciate experience, but it usually comes with ego

Jon

and more likely to be cooler to hang around with

Alex M.

letting go of what we know is something we do more and more difficult the more we learn

Evan

Lots of ego flexing going on in here today :P

Alex M.

15:38

and to fully understand and learn new things, you have to stop trying to apply the things you know with them

Jon

^

Almo

As a 43-year-old programmer who doesn't exhibit the behaviors listed above, I resent being lumped in with bad programmers who fail to learn new things.

Alex M.

it's okay almo

I know you're not like that and I do love you <3

Jon

We need a @Almo appreciation day

A day for programmers over 40 that have kept with the times, to be celebrated.

And as well, a reminder to the "others" to get out of their way.

MickLH

@Jon Could involve branding irons?

Old people love branding irons and cows n' shit

Jon

15:42

hmmm... that would be a tough sell

MickLH

I guess, most of the things I come up with are a tough sell

Alex M.

there has to be a situation here on chat where this would fit

Jon

lol, a guy wondering why his 1920x21600 image won't draw in monogame spritebatch

Icy Defiance

branding irons are okay, but those electric cattle prods are much more fun

John McDonald

@Jon heh

Icy Defiance

15:43

they don't need to be heated up first, so you don't have to give any advance warning

MickLH

I'd personally appreciate if everyone were fitted (from birth) with government sponsored shock collars that punish you for being objectively wrong or willfully ignorant

Alex M.

I'd hack it to punish you daily :D

Icy Defiance

or for avoiding NSA spying or saying anything bad about the president, any republican congressmen, or god

Kevin van der Velden

@AlexM. I'll keep it in mind for when the next random flag comes in on the bridge :)

MickLH

@IcyDefiance or for thinking of a way to escape the shock collar

Icy Defiance

15:44

right

Kevin van der Velden

@IcyDefiance but a nice hot branding iron can also be used as mood lighting

Icy Defiance

for a little while, but it requires recharging

MickLH

I think that's part of the nostalgia

It's slow and inconvenient

Icy Defiance

nostalgia is fine and all, but not many people can appreciate nostalgia without the inconveniences removed. the best is when you can make the audience aware of the inconvenience that used to exist, without actually forcing them to put up with that inconvenience.

MickLH

well damn, getting all programmer on that shit

don't you know people like to complain about being programmed even though they love it?

Alex M.

15:50

haha nsisfw: d24w6bsrhbeh9d.cloudfront.net/photo/av0DRw5_460sa.gif

MickLH

Nullsoft Install System Firmware?

Alex M.

pc master race

console peasants gtfo

Shotgun Ninja

lol

Transcript for