[ZenX2]

This code

SET A, 0x0000
SET B, [PC]
ADD B, 0x0002
SET [A] + 0x0100, [A] + 0xF030
ADD A, 0x0001
IFN A, 0x0005
SET PC, [B]
EXT 0x0000

writes 0-4 on the screen

This is really fun, aside from the infuriating problems

[Lexic]

ZenX2 posted:

This is really fun, aside from the infuriating problems

Sounds like all programming in general.

[ROBO_DONUT]

esalaka posted:

I can't see how having controls for such things would be more generic. Most games out there don't need a reticle with lock-on that also shows your missiles and energy, we're talking about a couple potential candidates.

Edited:

Point being implementing a window-based system is better than implementing absolutely fucking everything possible

You missed the point entirely.
I said generic. WoW-like. I'm suggesting an XML or JSON markup system with Lua or Javascript for scripting. Just a simple layout system with scripting.

Or, even leave the scripting out, since that can be separate. Just an absolute bare-minimum layout engine.


Bah, way off track. Ignore me.

[Nigey Nige]

Hey I heard you guys like programmer art



also mockups of games that don't exist yet and won't until my exams are over

shh this is totally programming

[Hexxeh]

Finally managed to get a stack trace out of Chrome:

GNU gdb (Gentoo 7.4 p1) 7.4 
Copyright (C) 2012 Free Software Foundation, Inc. 
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/ 
gpl.html> 
This is free software: you are free to change and redistribute it. 
There is NO WARRANTY, to the extent permitted by law.  Type "show 
copying" 
and "show warranty" for details. 
This GDB was configured as "--host=x86_64-pc-linux-gnu -- 
target=armv6zk-softfloat-linux-gnueabi". 
For bug reporting instructions, please see: 
<http://bugs.gentoo.org/>. 
(gdb) set solib-absolute-prefix /home/hexxeh/rootfs 
(gdb) file chrome/chrome 
Reading symbols from /home/hexxeh/chrome/chrome...done. 
(gdb) core core.chrome 
[New LWP 25926] 
Core was generated by `./chrome'. 
Program terminated with signal 11, Segmentation fault. 
#0  0x005e90ec in ?? () 
(gdb) bt 
#0  0x005e90ec in ?? () 
#1  0x3018b2f4 in __libc_csu_init () 
#2  0x40a154c4 in __libc_start_main () from /home/hexxeh/rootfs/lib/ 
libc.so.6 
#3  0x2a5edc24 in _start ()

Not exactly useful though. Anyone got any pointers?

[Catdaemon]

Hexxeh posted:

Finally managed to get a stack trace out of Chrome:

GNU gdb (Gentoo 7.4 p1) 7.4 
Copyright (C) 2012 Free Software Foundation, Inc. 
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/ 
gpl.html> 
This is free software: you are free to change and redistribute it. 
There is NO WARRANTY, to the extent permitted by law.  Type "show 
copying" 
and "show warranty" for details. 
This GDB was configured as "--host=x86_64-pc-linux-gnu -- 
target=armv6zk-softfloat-linux-gnueabi". 
For bug reporting instructions, please see: 
<http://bugs.gentoo.org/>. 
(gdb) set solib-absolute-prefix /home/hexxeh/rootfs 
(gdb) file chrome/chrome 
Reading symbols from /home/hexxeh/chrome/chrome...done. 
(gdb) core core.chrome 
[New LWP 25926] 
Core was generated by `./chrome'. 
Program terminated with signal 11, Segmentation fault. 
#0  0x005e90ec in ?? () 
(gdb) bt 
#0  0x005e90ec in ?? () 
#1  0x3018b2f4 in __libc_csu_init () 
#2  0x40a154c4 in __libc_start_main () from /home/hexxeh/rootfs/lib/ 
libc.so.6 
#3  0x2a5edc24 in _start ()

Not exactly useful though. Anyone got any pointers?

Segmentation fault can be a bunch of things, but usually means something to do with operations on invalid memory. Looks like you don't have any symbols from chrome loaded for the debugger to tell you what's actually going on.

[theJohn]

SFML is awesome :D

[DrLuke]

Hexxeh posted:

Not exactly useful though. Anyone got any pointers?

Sure:

0xbeef
0xf3ae
0xbc31

[Hexxeh]

Catdaemon posted:

Segmentation fault can be a bunch of things, but usually means something to do with operations on invalid memory. Looks like you don't have any symbols from chrome loaded for the debugger to tell you what's actually going on.

That's a build of Chrome with symbols included, it's like 1GB compared to the stripped version which is about 55MB.

[T3hGamerDK]

Hexxeh posted:

That's a build of Chrome with symbols included, it's like 1GB compared to the stripped version which is about 55MB.

Oh my, I feel sorry for you.

[ZenX2]

Absolute success, I added labels in under an hour

SET A, 0x0000
:loop
SET [A] + 0x0100, [A] + 0xF030
ADD A, 0x0001
IFN A, 0x0005
SET PC, &loop
EXT 0x0000

That does the same thing as the last code I posted :D

[chimitos]

Did I do Delta Time correctly?

//because Velocity = Speed = (1/2)*(acceleration^2), so a = sqrt(2*v)
acceleration = ((Math.sqrt(2*speed))*dt);

//then move according to the acceleration
y -= Math.round((acceleration*(Math.sin(degToRad*direction))));
x += Math.round((acceleration*(Math.cos(degToRad*direction))));

It seems to work, but considering how I was mainly guessing how it should go, It can't hurt to ask.

[ben1066]

Normals derived from the depth buffer suck, but I guess it's better than nothing.

[Lexic]

ben1066 posted:

Normals derived from the depth buffer suck, but I guess it's better than nothing.

Why aren't you using normals derived from the mesh triangles?

Edited:

triangle ( v1, v2, v3 )
edge1 = v2-v1
edge2 = v3-v1
triangle.normal = cross(edge1, edge2).normalize()

Edited:

Or since you're using source, the actual normal maps?

[ROBO_DONUT]

ben1066 posted:

Normals derived from the depth buffer suck, but I guess it's better than nothing.

You actually want to use normals derived from the depth buffer for some things. For example, in SSAO, using interpolated vertex normals results in horrible self-occlusion. Also, I use depth-derived normals for my anisotropic shadow filtering because it represents the actual geometry of the surface, rather than some arbitrary lighting normal.

But, yeah, they suck for lighting.

[ben1066]

Cause in alien swarm I don't have access to the base shaders, the best I can do is a PP shader using the depth buffer. That is until Valve decides to release the all elusive update with the shaders :P They are for SSAO anyway and I don't plan on touching the lighting. It's too easy to break.

[ROBO_DONUT]

You can't do SSAO if you can't apply it directly to the ambient term. If you apply to everything as a postprocess it looks awful.

[Naarkie]

I have made a game for Ludum Dare and you should play it. Its a minimalistic roguelike.
Yes its very bad.

Doonlaod
Read the readme.txt to find out about Lewis' marvellous super-powers.

[ben1066]

ROBO_DONUT posted:

You can't do SSAO if you can't apply it directly to the ambient term. If you apply to everything as a postprocess it looks awful.

Really? I would have thought it would have looked better than no SSAO.

[esalaka]

ROBO_DONUT posted:

You can't do SSAO if you can't apply it directly to the ambient term. If you apply to everything as a postprocess it looks awful.

So basically you have to go deferred if you want to SSAO?

[ROBO_DONUT]

ben1066 posted:

Really? I would have thought it would have looked better than no SSAO.

SSAO is an approximation of the visibility function for indirect light. If you apply it to direct, too, it just looks like a cheap drop shadow.

Here's an example from some random dude's XNA blog, where he's clearly doing it wrong: http://i.imgur.com/ktQfh.jpg

Of course, you could always do it anyway, call it 'NPR', and nobody will be able to argue with you.

esalaka posted:

So basically you have to go deferred if you want to SSAO?

No. Deferred implies that you render your geometry to a g-buffer in a single pass, then do shading from that g-buffer.
This isn't strictly necessary for SSAO. As long as you have access to the depth information, you can do SSAO as part of the ambient pass. Forward, deferred, or otherwise.

[esalaka]

ROBO_DONUT posted:

No. Deferred implies that you render your geometry to a g-buffer in a single pass, then do shading from that g-buffer.
This isn't strictly necessary for SSAO. As long as you have access to the depth information, you can do SSAO as part of the ambient pass. Forward, deferred, or otherwise.

wasn't entirely sure of the implications of the term

[raccoon12]

I just can't wrap my head around assembly... it's so convoluted


edit:
those dumbs really hurt ;(


also, I understand how it works.. it's just annoying complex

[ROBO_DONUT]

It's only convoluted if you're learning on a CISC system.

Honestly, I'd call assembly in general anything but convoluted. Maybe "painfully simple" in the sense that it's difficult to do anything because you're only working with the most fundamental operations.

[Darkwater124]

Teh end.

Now it's time to make the fancy stuff.

[Krizzu]

Darkwater124 posted:

Teh end.

Now it's time to make the fancy stuff.

How long are you coding in L2D?
And how Can you compile .love to .exe? Cuz when Im using cmd>copy /b blah blah, it says it's not valid win32 application.

[Eric95]

DINO-CAB is done!



Play dat shit here: http://www.ludumdare.com/compo/ludum...eview&uid=3430

[Mordi]

Finally (the not-quite-full range of ASCII characters).

[Darkwater124]

Uhh......

Anybody wants a LOVE-targeted GLSL script for a round HUD? (I was trying to make a radial/zoom blur)

Krizzu posted:

How long are you coding in L2D?
And how Can you compile .love to .exe? Cuz when Im using cmd>copy /b blah blah, it says it's not valid win32 application.

First off the official name is LÖVE, but I've been using it since Garry came with those programming assingments, a few months ago.
And you can't really "compile" a .love to an .exe. You can, however, package love.exe with your .love, meaning you can still see the contents with WinRAR or the like. If it says it's not a valin Win32 app, you must have done something wrong as I use that command too, and it works just fine.

[origamiguy]

I just made the first ever player-made EDGE level.

[Mr. Smartass]

Finally got layered saving working.
How can I make my code net-friendly before it's too late?

[ace13]

Finished my first ever Ludum Dare entry, Tiny Hack/.

As the name suggests, it's a small hack and slash complete with silly references (Not nearly enough of them though). It's probably riddled with bugs and all that, but I'm sort of proud anyway.

Behold my programmer art!

[origamiguy]

Little experiment with FoV triggers:

[Ac!dL3ak]

Mr. Smartass posted:

Finally got layered saving working.
How can I make my code net-friendly before it's too late?

that looks nice

[D33Jay]

my god the end result of this game was terrible. need to think of a simpler game next time.

[supersnail11]

I came up with a great idea for a game. Basically, you'd have to beat the computer with pathfinding. Try to get the best time. Of course, you'd have to dumb down the computer, else it would be really hard to win.

The game name should be "When you wish upon A*"

[Darwin226]

supersnail11 posted:

I came up with a great idea for a game. Basically, you'd have to beat the computer with pathfinding. Try to get the best time. Of course, you'd have to dumb down the computer, else it would be really hard to win.

The game name should be "When you wish upon A*"

A classic example of "Oh, that's a great name I just came up with. Time to build a game around it."

[ZenX2]

I'm making my assembler more forgiving of whitespace and stuff
Also comments

Edited:

Ok now I wrote some preprocessor thing I guess

[false prophet]

I've been playing with javascript as a game platform. It's pretty horrible.
I'm trying to get javascript embed into a C application but it's not going to well.

[swift and shift]

Making Ruby faster:





I have the foulest code ever that converts Ruby bytecode into C, compiles it and links it in to Ruby.

Edited:

and this is the C my thingy generates: