Game Development - 2013-02-15

« first day (925 days earlier) ← previous day next day → last day (4107 days later) »

1:23 AM

Correct me if I am wrong but intent is really important in regards to determining if something is a copyright infringement.

so wouldn't making this post http://gamedev.stackexchange.com/questions/49390/game-loosely-based-on-a-copyrighted-story alone put you in a bad spot?

@ClassicThunder the point is lawyers don't consider your intent, they just consider winning the court.

Yeah I know but creating a character identical to an existing one is not a violation in and of its self. It has to be shown there was intent to copy the existing one if I understand correctly. At least this is what I understand.

2 hours later…

3:25 AM

0

Q: Opengl normals not working

I have a model rendered in LWJGL. I have calculated and set correct normals and set up a light. However, the model is lit incorrectly. One half of the model is its normal color (magenta), and the other half is white. Both of these parts are shaded flat. Does anybody know why and how to fix this?

opengl java lighting lwjgl normals

Is it horrid of me to want to post an answer stating "This is caused by a bug in your code. Fixing the bug will cause the problem will go away." ?

it could be a bad model too

The question says that the correct normals are set. Therefore, that is un possible. ;)

4:15 AM

(More seriously, yeah, that's completely possible/likely)

well, the fact that it's half-half I suspect is the model being converted with a mirror modifier in the editor, thus flipping half the normals. the fact that it's flat-shaded magenta, well, fixing the buggy code should fix that problem

4:48 AM

Honestly, my suspicion is that the normals aren't unit length.

If your normals are, like, 1000 units long, you wind up with this amazing overbrightness, with one side clamped out at pure white, with the other side only affected by ambient lighting. Easy to get the described effect that way.

5:12 AM

oh I see. you mean half as in, the away-facing faces

I interpreted half as in, the entire left side of the model

1 hour later…

6:13 AM

if i use

b2PolygonShape shape;

shape.SetAsEdge(b2Vec2(screenSize.width/PTM_RATIO*0.10 ,screenSize.height/PTM_RATIO*0.80), b2Vec2(screenSize.width/PTM_RATIO*0.10 ,screenSize.height/PTM_RATIO*0.35));
body->CreateFixture(&shape, 0.0f);

der ll be edge, but will be visible only when world->DrawDebugData(); is active...

can i relate setAsEdge with sprite , or how to make visible edges when drawDebug is disable??

6:37 AM

Oh cool, I got upvoted on both of my answers from today.

Best part is, I'm far from being a professional game developer ;P

@BaSha that's the only way, unless you call that Debugger yourself.

it's fun, an answer of mine which received 6 down-votes is now only -2. I think the culture of this site is changing!

7:10 AM

@Chunk-e-Yamani heh, which one?

2 hours later…

8:43 AM

-2

A: C++ GameState management

The most common approch to GameState algorithm is to have some game states and program can freely choose to run any one of them, this means there is a GameDirector which has a pointer to current running state, and in each update cycle it'll call 'currentState->run()`. in this approch GameDirector...

2 hours later…

10:52 AM

the downvoters appear to be correct

there's no need for all that

a game state can be something as simple as if..else/switch on an ID variable

the culture of gd.se is this: the first impression is usually more or less correct, but with time, OCD/OOP people come along and they mess up the ratings

there's lots of OOP in your case, which explains the upvoting of your answer, as well as the downvoting of my answer, which was very much hating on OOP

and the "good practices" and whatever

it also explains why the communist duck has an upvoted useless comment

1 hour later…

12:22 PM

@snake5 actually I didn't really mean anything like that. I just replied the architecture I've previously seen in other game engines.

12:44 PM

I guess you've all noticed by now but anyway, photoshop v1's source code is released!

computerhistory.org/atchm/adobe-photoshop-source-code

it's a combination of r,p,t,a and inc files!

1:10 PM

can i still vote or downvote a closed question ?

1:32 PM

@Chunk-e-Yamani it's just that the question wasn't about that

@snake5 he asked how to handle game states, and I explained how I've seen other engines handle it, how is that irrelevant?

@Chunk-e-Yamani he asked about game states

you just said "scenes" :P

@snake5 typo

and most of what you described in the code was irrelevant to states

render managers, crap like that

@snake5 so you mean I just have to remove 9 lines from that code?

1:38 PM

9? I see a lot more

I'd say the number of relevant lines is 9

or somewhere around that

2 hours later…

3:24 PM

hahahah textfromxcode.com

3:38 PM

@Chunk-e-Yamani I had not noticed; thanks for the link!

ToddersLegrande

3:54 PM

I might have purchased the first season of DBZ last night....

were you drunk or something?

ToddersLegrande

lol

Hardly. I just really wanted to watch it.

Ah, ok. :) That's a good reason.

ToddersLegrande

I kind of love how I can watch videos I purchased on the Xbox and the PC!

Probably my phone too if I had to guess...

:)

ToddersLegrande

4:03 PM

Last time... on DRAGON BALL Z!!!!

4:40 PM

:D

5:17 PM

They say that PS4 will have something GPU related that has "dual shader engines" does anyone have a clue what is that? is it something they found in Roswell?

I don't. I have a little info on PS4 from someone working with one of the dev kits, but I don't know about the shader thing

5:30 PM

@Kikaimaru sounds crazy

if anything

modern GPUs already have many unified shader cores and "engine" is such a wrong word to be used here

marketing talk or some journalist made it up, I'd say

google can find the phrase around various forums, which are first in the results list

so the new version is that some guy just made that stuff up

ah!

vgleaks.com/world-exclusive-orbis-unveiled

journalist made it up, after all

they like to do such crap

like "web OS"

Bartek Banachewicz

5:54 PM

PS4 is gonna have ATI shit

who's ATI? :p

lol forgot they renamed themselves to just AMD

6:10 PM

AMD bought ATI

and it's not "ATI shit" since, clearly, it's not shit :D

I've only had terrible experiences with ATI cards, but I haven't bought one in 8 years or so so maybe AMD has improved them

I had a 5770 some time ago, I don't remember what but I had some trouble with it...went back to nvidia

6:29 PM

@Jimmy 8 years? well, definitely has

back then the Linux drivers were so bad

I've only had terrible experiences with nVidia cards, but they were all cheap/integrated or from questionable manufacturers

I'd fry a graphics card every 6 months or something

I've had bad experience with NVIDIA Quadros

but all the Geforce cards I've had were great

ToddersLegrande

I've only owned AMD / ATI cards since about 2004 or so. I've been pretty happy

I can really only think of one or two games I played where the ATI drivers would freak out if you turned on some advanced features. And they were eventually patched

Bartek Banachewicz

7:15 PM

Whatever, nobody buys AMD/ATI cards or processors anymore

they dont?

aren't radeons doing well these days?

although it does seem Intel and ARM are having CPU teatime and AMD wasn't invited to the party

@BartekBanachewicz I do

I bought an ATI for my mac. Radeon 9700 i think. was great

you defiled the sanctified body of a mac with aftermarket hardware? :o

7:26 PM

Warcraft III was running way too slow

@Almo ATI was great those days,

would upgrade my MacPro's card, but I can't get one for a mac that old any more

it just sucked around hd1000 through HD3000 series.

and then it became great again

Bartek Banachewicz

@snake5 well, my condolences

1

A: glsl vector plus scalar

http://www.opengl.org/registry/doc/GLSLangSpec.3.30.6.pdf From 5.9 The arithmetic binary operators add (+), subtract (-), multiply (*), and divide (/) operate on integer and floating-point scalars, vectors, and matrices. (...) One operand is a scalar, and the other is a vector or mat...

^ that's how good ATI ~~cards~~ drivers are

@BartekBanachewicz nVidia is worse

Bartek Banachewicz

7:36 PM

@snake5 that's a joke, right?

they allow all kinds of crap in their opengl drivers

Bartek Banachewicz

oh, yeah, sure.

and they can't even produce proper error messages in the debug log

@snake5 like opengl? :p

@melak47 yeah, that too :D

7:36 PM

zing :p

Bartek Banachewicz

@snake5 that's too bad you can't write proper GLSL :)

nVidia allows HLSL in GLSL, I've heard :D

Bartek Banachewicz

oh no.

what happens now?

makes it kind of hard to develop something there

Bartek Banachewicz

@snake5 that's utter bullshit

7:37 PM

won't notice a thing until you run it on ATI

@BartekBanachewicz you're hateful and I'm not sure why

Bartek Banachewicz

@snake5 because what you are saying doesn't make sense whatsoever

It's not a hate, it's a fact.

If you write non-standard code, it's your problem.

oh, "non-standard code"

that statement was full of shit

Bartek Banachewicz

Standard-compliant code compiles on NVidia, that's all I want to know.

@snake5 excuse me?

Curious: what standard deals with the shader languages?

I don't know a lot about core rendering stuff

Bartek Banachewicz

@Almo GLSL specification

7:39 PM

interesting: will go look that up out of interest

it doesn't look like you've ever developed anything bigger than a "hello world" app and you're now saying that nVidia allowing non-standard things isn't bad

Bartek Banachewicz

@Almo that's my everyday essential tool in job :)

:D

surely when ATI does something like that, it is bad, right?

Bartek Banachewicz

@snake5 I work in Intel in Graphic Drivers team.

Is it bigger than hello world for you?

7:40 PM

@snake5 very. =)

depends :D

oh, intern :D

@snake I agree with @Bartek. having "extra" features isn't a problem. Not being compliant with the standard is.

that kind of changes things

So OpenGL is still under active development

Bartek Banachewicz

@Almo of course it is.

7:41 PM

@Jimmy no, it is a very big problem

good to know

Bartek Banachewicz

@Almo ever heard about OpenGL ES or WebGL? ;)

when you translate code, HLSL to GLSL, you want to know that nothing's missed

A wise programmer once told me, "There's nothing direct about DirectX"

Bartek Banachewicz

@snake5 if you can't write proper code, yes.

7:42 PM

so... you know nothing is missed when "it compiles"?

@Jimmy that's pretty much the only option with opengl

@BartekBanachewicz ooh

I dunno, there is a spec and everything...

Bartek Banachewicz

@Jimmy Real codehz don't use specs or references, don't you know?

@Jimmy feel free to check every shader against 100-something pages of spec. text

Bartek Banachewicz

@snake5 Compilation errors are also in the spec.

7:44 PM

the problem isn't when shaders don't compile

it's when they do compile, even though they shouldn't

Bartek Banachewicz

@snake5 that means you wrote them badly, right?

no

it means that the crappy driver lets too much stuff flow through freely

Bartek Banachewicz

Well, sorry, but are you implying that driver should be responsible for your bad code?

I'm also pretty sure that there's option to enable strict compilation, i.e. without NV implicit extensions.

Because, as I said earlier, GLSL spec also says when something shouldn't compile.

yeah? where's that? some special NV extension function? wow

Bartek Banachewicz

I don't get you.

You write non-standard code, and whine that it compiles.

That statement is funny by itself.

Under no circumstances should you check if the code meets standards by compiling it.

If you want to check if the code meets standards, check in standard. Simple as that.

7:48 PM

oh, if OpenGL did have some sort of external compiler, I wouldn't have to do that

I mean, I admit it's not a good system when something implements looser semantics than what is required. This is how bad programmers learn bad practices. (case in point, IE6)

@BartekBanachewicz you're funny

but at least it's not a "bug".

I had an interview question where they showed me some C++ code that compiled, and asked if there were anything wrong with it. I said it looked odd on one line, but that I didn't really know what it would do.

Bartek Banachewicz

1 min ago, by Bartek Banachewicz

Under no circumstances should you check if the code meets standards by compiling it.

7:50 PM

He was happy with the answer I gave, because I spotted the weird line. It was something that would compile, but had "undefined behavior," and thus shouldn't be used.

Bartek Banachewicz

Good point, @Almo.

@BartekBanachewicz I don't really care about standards, I just want to know that my code won't compile only on nVidia

Just because something compiles doesn't mean it's right, even in something like C++

2

Bartek Banachewicz

@snake5 y'know, that's what standards are for

same goes with ATI, if/when they pull such shit

7:50 PM

+++++ almo

like, just because it compiles, you'd still have to test if everything is shaded purple on some particular manufacturer's card

@BartekBanachewicz so would you check each shader for me, anytime? because I clearly don't have the time and precision and energy to do that manually

Bartek Banachewicz

@snake5 I am checking my shaders myself and I see no reason for you not to. You are just seeking excuses for being lazy.

@Almo how about "especially in something like C++" ;)

:D

@BartekBanachewicz yeah, right.

Bartek Banachewicz

7:53 PM

@snake5 I mean, I wouldn't hire somebody like you.

@BartekBanachewicz it's not like you could

lets keep discussion civil, everyone

^

Bartek Banachewicz

okay.

@snake5 that's not the point. All I'm saying is that depending on a particular compiler is bad, regardless. What if it has bugs?

Recently guys on SO found a bug in MSVS 2012 compiler.

Yeah that was a hell of a thing.

I sent that around to the programmers where I work

7:55 PM

@BartekBanachewicz it certainly has. but I still need an automated tool to check the shader

it's much easier with Direct3D where there's only one compiler

what bug? the thing where the compiler crashes? :)

@BartekBanachewicz that's not new, there are way too many bugs to find that compiler.

Bartek Banachewicz

@Chunk-e-Yamani that's why I ain't using it.

@snake5 what if that tool has bugs?

@BartekBanachewicz pff, you scared? :p

@BartekBanachewicz then someone will fix them

7:57 PM

@BartekBanachewicz still I prefer it, when my alternative is gcc.

Bartek Banachewicz

@Chunk-e-Yamani I have free ICC

but I won't have to re-read the standard all over again just to fix a shader

Bartek Banachewicz

@snake5 you remember most important parts quite fast; personal experience :)

And more and more you work with it, less you have too look into spec.

Y'know, spec isn't written for robots; it's pretty intuitive, actually.

@BartekBanachewicz of course, but bugs tend to sneak in either way, especially in conversions

Bartek Banachewicz

@snake5 that's mostly because people write code that compiles just for them

And we have "code works on my machine, but fails everywhere else". Now what?

Process of fixing it is painful.

And essentially boils down to writing good code in the first place.

8:00 PM

I hate it more when it works on everybody else's machine :/

hahaha

@BartekBanachewicz sure it does, and I get that, but it's not as much about writing, as it is about some tiny difference that doesn't get noticed in conversions

like vec4 creation functions

Bartek Banachewicz

These things change from version to version.

what kind of arguments they allow, in what order, what maps to what etc.

Bartek Banachewicz

@snake5 do only explicit conversions then. Problem solved.

8:03 PM

those explicit conversions are still different

nothing's solved there

Bartek Banachewicz

vec4(float(1)) is pretty descriptive, don't you think?

At least the intentions are clear.

quite

Bartek Banachewicz

Still; everything you aren't sure if is allowed, checking in the standard takes a few seconds.

While working at Intel I actually started writing shaders that work from first run. I just knew what I was writing.

I think that's a very good skill to have.

did they work everywhere? :D

Bartek Banachewicz

@snake5 everywhere I checked :) On iPad, tegra tablets, ARM reference implementation, and, of course our HW.

So pretty much yeah.

8:07 PM

cool.

they they compile on a train? did they compile in the rain? did they compile with a fox? did they compile in a box?

3

hahahaha

Bartek Banachewicz

lol.

I thought I had that with C - writing things that worked from the first run. surprisingly though, it wasn't quite true

Bartek Banachewicz

C is painful.

8:08 PM

things tend to get missed in complex systems

Bartek Banachewicz

too many pointers

shaders aren't as complex but they have their tricks

Bartek Banachewicz

well, anyway, the discussion was nice, but it kept me from writing code, actually :)

the problems with new/old functions caught me more than once

texture/texture2D and such :D

Bartek Banachewicz

I'm going back to my QuadTree, see you guys around. Or look for me in the SO Lounge<C++>.

8:10 PM

yeah, me too

got a scripting language to work on :D

3 hours later…

11:36 PM

Any one know websites where to find people interested in working on open source game projects?

« first day (925 days earlier) ← previous day next day → last day (4107 days later) »

Transcript for

Feb '1315

Game Development

Game development and other polite discussion. Game development...

game-design game-industry rubber-ducks

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile