hey guys ... does anyone know where I can find articles on getting started with native dx ?

directxtutorial.com was somewhat ok, I think

ok thx

anyway, I'm curious about something myself... does anyone know any path triangulation algorithms?

I wrote one but surely I can't be the only one to do so

path triangulation ?

you mean path finding ?

like A*

no, I mean vector graphics paths

oh

then no im a retard and too stupid to even understand that question

for simplicity, just lines, without all the curves

seems to be much more simple than what this guy wrote here: cosy.sbg.ac.at/~held/projects/triang/triang.html

trapezoidation from several other algorithms isn't close to my approach either

I constantly find myself thinking that a Megahertz is actually 1024 kilohertz.

Nice

@snake5 Last time two times I implemented this, I completely avoided complex triangulation

complex being self-intersecting polygons?

Instead I computed a convex hull and the crude version would use the stencil buffer to compute XOR crossing

oh

Then my improved version generated a formula to compute the pixel value by 2D raytracing

@AidanMueller 1024 KiloHertz = 1 MebiHertz

I was making this for mesh generation from svg files

I did mine for SVG rendering, because my engine supports SVG textures

The raytracing approach allows to directly render the bezier curves, instead of approximating them by polygons

so while raytracing isn't really out as an option for me, I'd like to avoid it since I can't tell at what resolution it would be rendered

Looks fucking sexy with a little AA

yeah, it does

I'm going for the GPU-accelerated route, though

anyway, just wondering if someone in the whole world has written an algorithm just like mine

lol, so did I

I wrote an algorithm like yours during evaluation of the different strategies I came up with

how did it work?

But I threw it away because the ray-tracing strategy beat it up to cubic

It works like I explained lol, compute a convex hull, then convert each path segment into an equation

ah, that

I wrote a completely different one

I understand that, I wrote that too first

without hulls and equations

You asked if anyone has done similar :P

I skipped straight to the good part

Although I found it unreasonable to go past cubic bezier with this approach

I'm looking for one that's as close as possible to the one I wrote

My code converts higher order beziers into a group of lines, or lower order beziers depending on a "goodness" factor

That part is resolution dependent

yeah, I hadn't got to that part yet, just did the self-intersecting polygon triangulation part

I think microsoft made a good paper on this topic

Check out how they did SVG rendering in the latest IE

It will fit into your idea pretty easily

where can I find info about it?

google, bbl told my girl I'd come eat the sandwich she's making me lol

:D

I am googling

no luck so far

Quiddich is some seriously bad game design.

Yes

I forget... is there a way to win without catching the snitch?

Or is the snitch the only way to signify the end of the game?

you can get points for doing the ball through those three rings, I think

but the snitch will end the game outright.

nobody had every caught it and won that way till harry did

so... shoudn't he have been doing something more useful with his time on field if it had never been caught?

anyway, it's all a bit fuzzy from the movie. I just remember being shocked by the design, and being bored watching them play it because of this.

to be sure, asking a good game design from a narrative author doesn't make a lot of sense.

she would have been better off being vague about the rules, I think.

@Almo it was not that bad

it was kinda basketball, meaning you got points for scoring through those rings

but getting the snitch was worth something like 15 goals

so if you were behind more than 15 goals, you would probably not want to catch it

but no one had ever caught it before

who said that,

almo did...

it was the only way they could have ended the game

there was no time limit

I don't know enough potter stuff, but it's what I remember from the movie.

no time limit?

harry was just exceptional in how fast he could find and get it

great game design there.

like there's a board game of soccer with a time limit.

Sounds subjective

well, that's how I remember it.

no rules for determining if one player is stalling for time.

get ahead, suddenly start taking more time to think about your moves

Beat them with the paddle if they are stalling for time

I remember in one point it was said that a game last more than two days because they couldn't catch the snich

:D

you're not making it sound like the game design is workable :)

wikipedia agrees with Ali: "The seeker who catches the Snitch scores 150 points, and only the capture of the Snitch will end the game. It is mentioned in the books that on rare occasion, quidditch games have been known to last for months, so it is of key importance to catch the Golden Snitch as quickly as possible."

@Almo stalling doesn't really make sense if there is no time limit

but at the same time, that makes everything in the game worthless except for catching the snitch

so definitely horrible game design

so when someone gets the snitch, the game ends, and you get 150 for it.

@IcyDefiance I didn't remember the months part!

but if you are not within 150 of the win, you wouldn't try to catch it

You would instead prevent the other team from catching it

sounds like this game was designed for "fantasicness"

not for playability

not only you can't try to catch it, you should also prevent enemy catcher from getting it

which makes sense given its source, a novel.

Thanks for clarifying; apparently the movie was unclear about it. But I still think it's bad game design. :)

as far as I can tell, the game is completely practical, and maybe even fun to play/watch

of course if the longest games end in less than 1-2 hours

on the merits of shipping first on iOS or android. Each gives a different opinion.

if it was up to me (and I had OSx available) , I would have gone for iOS first

simply because it's way easier to test your app against available devices

Apparently there are colleges that play / compete in which the snitch is only worth 30 points. Fascinating.

that's one of the big reasons the iOS proponent gives.

testing

time to go see Lewis Black live

see you guys later :)

have fun!

Just solved a problem I've been working on for 3 weeks!

Yay!

Go you!

Time to make some placeholder models...

This looks pretty efficient to me.