pouët.net

So, Grid.org has finished all projects.

How do we earn glöps now?

that topic title scared me.

Admit it - you're done. Nothing to do anymore in your life. :)

I came in here expecting to see a post from Gargaj about... well, anyway, nothing to see here, moving along...

we could switch to folding@home or any of the other 10 thousand millions projects.

Actually...

If someone would take up one of my projects that I'm not gonna get to anytime soon and create a version of Electric Sheep (electricsheep.org) that is scene-related (read: I don't care what it renders so long as it's sceners controlling it and it produces something we make) I'd swap my machines from electric sheep to it. I'd also host the controlling servers, provide stats, etc.

I've some experience in producing the stats tracking and graphs, so that portion isn't that hard; it's the 'distributed iterated fractal' that I haven't fucked with yet.

I don't care if it's a fractal bit, distributed raytracing, or generating frames for something else. So long as it's visual (and optionally aural) and can be distributed, I'm willing to donate some server back end + stats time to it.

Anyway, if anyone -is- interested, get ahold of me.

Ok, so they shut down grid.org. But did they ever show any results coming out of it?

I'd be careful about saying things like that...

"WTF? You mean that for the last 6 weeks my computer has been rendering a 30-minute cutscene for Metalvotze's long-awaited epic Presswurst III?" :-)

@gasman
we have ray tracers which can beat your ray tracers
(so no need for grid computing pw 4 ;)

seti@home ;)

scene@home :)

perhaps we can use the idle processing power of sceners to calculate worthwhile demos.

they've been resurrected.... it seems link me sexxy

Worthwhile demos run in realtime.

anyone succeeded in recovering his old account at grid.org ?
or do we have to create a new one ?

raytracing would be interesting (even if we have raytracers that beat their raytracers, still 15 million tris/second is nothing usable for any nice looking image, not even with recent fast AO tricks, so such a backend could be usefull to do the images worth a raycaster). However I certainly find fractals even more interesting. Or even, we could arrange something, win the http://www.eternityii.com/, become rich, and program demos the rest of our lifes...

Are there many cases where a huge amount of pre-calc would be handy for an effect, but where it wouldn't produce gigs of data?

ooh, lets make all the machines in the world generate every possible combination of bytes possible for a 4k intro, run it and test it. we'll get the perfect one someday!*







(*im not serious)

yeti@home :)))

ohhhhhhhhhh, i know you're not serious, but... that would be so interesting.... if 256^4096 was not such a big number (10^9864). I'm stunned, even for a 16 byte intros (10^38 intros) it would be completelly impractical in brute force.

toxie, I meant 15 million RAYS/second per second of course :)

iq: not every combination of 16 bytes would run, and most would write nothing to the screen. Actually, it might be practical if you had a way of trying every combination that rand and wrote to the screen.

or use a 32 bits random seed to generate an expression evaluation abstract tree, which can me interpreted or compiled in JIT.. by tweaking the seed i'm pretty sure you can get interresting results that are worthwhile watching.. it's not so far from what i do to parametrize routines with lots of parameters..

one of the 4k intros will actually format your hard disk...