next »

[Dr. Cooldude]

http://www.medicaldaily.com/articles/9496/20120411/alcohol-solving-skills-analytical-thinking-creativity-study.htm

I think this is mandatory

Haha! That made my morning! Thank you sir.

[Graham-]

Programming is definitely _not_ a task that doesn't depend on focus.

[Muz]

Programming is definitely _not_ a task that doesn't depend on focus.

If you're one of those people who like to go "what do you think of my idea/drawing" for a few months, you'd probably be more productive with a little alcohol.

[Graham-]

Hunh? I don't do that. What are you talking about....

The double negative reverts to: programming takes focus, hence the alcohol benefit is situational. You have confused me sir.

[st33d]

I can draw and program. I can draw blind drunk. If I'm programming I end up making a mess that I can't figure out when sober because I can't remember what I was busy breaking.

[Graham-]

Programming drunk is really a whole different thing, because your mental tray of shit randomly resorts and wipes.

You just need a really focused work-flow system. Then you can drink.  .... yeah right.

[Fallsburg]

I think it's because it brings me back to college, but I'm pretty darn good at implementing algorithms while a little drunk.  It helps me turn off the part of my brain that gets bored by just churning out code, and I think I tend to make fewer mistakes/work more fluidly (i.e. not checking random websites as much).

[ThemsAllTook]

I don't drink enough for it to really affect my coding, though I think I feel a little bit less motivated to do stuff after a beer. I have observed that I do different types of coding tasks better at certain times of day. Mornings are great for thinking and solving complex problems where I'm not typing out a huge amount of code; evenings are better for just hammering out large amounts of stuff without thinking too hard about it.

[Salsario]

AFter many times trying to create a game but stopping halfway thru I can finally report I am getting somewhere with a project. It isn't a game however. But a leveleditor and game-engine, xna in winforms. Sort of proud that I finally got that part working. It also forces me to think in advance what my game will need, allthough when the basic functions are all implemented (sometime this week) I think I will just start to make some fast basic games ala tetris, pong, sokoban, to get me going on the enginepart.

Who knows, I might actually finish a game this time around. And if not, I still got a nice leveleditor :D  (and I do drink while coding )

[Muz]

Hunh? I don't do that. What are you talking about....

The double negative reverts to: programming takes focus, hence the alcohol benefit is situational. You have confused me sir.

I don't mean 'you' as in the 'person I quoted' 'you' but you as in the 'reader of this post' 'you'

Point was that some people program better drunk if they're the type who spend a hell lot of time thinking about how to do something or procrastinating and worrying rather than just doing it. If you've planned everything out in advance, it's usually easy to just follow it.

Alcohol is liquid confidence because it suppresses the smart part of your brain which instills doubt in your abilities.

Programming drunk wouldn't be too much difference than programming late at night, which a lot of people do. But you wake up with programmer's hangover, i.e. spending a week fixing bugs introduced in that 3 hour burst of productivity. But it's still a step ahead from doing nothing!

[Graham-]

ohhhh....

I see.

alcohol removes analysis-paralysis.

I don't drink enough for it to really affect my coding, though I think I feel a little bit less motivated to do stuff after a beer. I have observed that I do different types of coding tasks better at certain times of day. Mornings are great for thinking and solving complex problems where I'm not typing out a huge amount of code; evenings are better for just hammering out large amounts of stuff without thinking too hard about it.

Me too.

[InfiniteStateMachine]

Got 3 different types of VBO rendering pipelines working in OGL.

Interestingly indexed rendering will actually run slower on GPU's that dont support it (at least from what I found on the internet). It seems it will do the indexing on the CPU instead. Odd because I'm used to a hard crash if something isn't supported.

Now trying to find out if indexing support is detectable in OGL.


Also interleaved arrays or VBO's becomes quite noticable when rendering over 20-30 million triangles.

[ThemsAllTook]

Interestingly indexed rendering will actually run slower on GPU's that dont support it (at least from what I found on the internet). It seems it will do the indexing on the CPU instead. Odd because I'm used to a hard crash if something isn't supported.

Huh, hadn't heard about that. When you say indexing, you're talking about glDrawElements(), right?

Also interleaved arrays or VBO's becomes quite noticable when rendering over 20-30 million triangles.

Noticeable as in significantly faster than non-interleaved?

[InfiniteStateMachine]

Interestingly indexed rendering will actually run slower on GPU's that dont support it (at least from what I found on the internet). It seems it will do the indexing on the CPU instead. Odd because I'm used to a hard crash if something isn't supported.

Huh, hadn't heard about that. When you say indexing, you're talking about glDrawElements(), right?

Also interleaved arrays or VBO's becomes quite noticable when rendering over 20-30 million triangles.

Noticeable as in significantly faster than non-interleaved?

Yeah I mean using glDrawElements with an array of vertex indexes. I was pretty surprised. Mind you this explanation is just from me googling things like "glDrawArrays faster than glDrawElements" and finding that answer in some forum thread. My test machine has some onboard video. It was about a 20-30% drop in performance with indexing enabled.

The interleaved array was about 10% faster, I guess I was exaggerating when i said "quite noticeable". Nothing too signifigant. Also only noticeable with 10's of millions of triangles on-screen. Otherwise there was no noticeable difference.

[Average Software]

The interleaved array was about 10% faster, I guess I was exaggerating when i said "quite noticeable". Nothing too signifigant.

A 10% performance impact is extremely significant in any context I've ever worked in.

[InfiniteStateMachine]

Yeah I guess I just didn't want to oversell. It certainly is significant. It's certainly a great boost for maybe an hour worth of work. Probably would have gone a bit faster but I misunderstood the meaning of stride when dealing with interleaved arrays.

In the interest of sharing. If you had 3 vertices (floats) and 4 color values (floats) you would use stride and offset like so:

format being VVVCCCC (3 vertex floats, 4 color floats)

glVertexPointer(3, GL_FLOAT, /* stride * */(3 + 4) * 4, /* offset */0);
glColorPointer(4, GL_FLOAT, /* stride * */(3 + 4) * 4, /* offset */3 * 4);

so stride would be the number of elements from the first value of the vertex or color value to the next first value of the vertex or color value. The offset would where the start of the color data starts. Since the vertex is first it has no offset.

The 4 represents 4 bytes (sizeof(float));

[_Tommo_]

I fixed a very ugly bug due to a "state leak" in DirectX, which caused random stuff to become black depending on the rendering order... and this led me to fix 3 other bugs in OpenGL using the same solution in the other render
I wish this would happen for every bugfix

[PompiPompi]

Aint this lovely :D

[Muz]

Some guy designed something in Photoshop. Fucking Photoshop and asked me to recreate it in Android, with the hell of converting a cramped Photoshopped design to something half its resolution.

Spent the whole day on it.

Showed it to them, and they were like "whoa"

It's an awesome compliment when someone with high standards goes "whoa." It's better than if a really attractive girl says "I want to have sex with you now." I get a "good job" and "it looks good" every now and then, but nothing beats "whoa"

It's great how smartphones have really vibrant displays and actually look better than it does on any PC monitor. It feels good creating something prettier than Photoshop. To top it off, it actually runs really smoothly on a low end device.

[R.D.]

Probalty because smartphones are made by people you think about what they are doing. Even animations are much smoother on a smartphone then on any high-end pc...