Random "work in progress" shots
category: general [glöplog]
some difficulties we had to overcome while making this:
- not having enough time( yeah, what else is new? )
- RenderMonkey crashes when reaching the instruction limit (anyone still has the email adress of the ATi-schnitte? )
- coding an own tool in a couple of days.
- writng shadercode, avoiding hitting the actual instruction limit
- not getting the compile to inline the party you want and not inilinig the party you don't want
party = parts.
(don't ask.)
(don't ask.)
abductee:
the rendermonkey-project is obsolete since 2007 or sth...so contacting them will yield in no help i guess ;)
just have an eye on the bottom line, it tells you about the ALU-count etc.
so you just have to know about what instruction takes what amount of ALU-cycles and you should be fine! (traditional democoding! YAY!)
my problem with rendermonkey is the exeeding of registers...its a mistery to me how they solved that, if a shader is working in RM it is possible it is NOT working in your c++-counterpart sometimes, due to the ShaderCompilers seem to be different between RM and DirectX! this disqualified the most-best scenes i ever had and i still hate having gone the route of obsolete RenderMonkey!
in other words: dont trust in obsolete 3rdParty-Software!
also: making a tool for 4k is overEngineering, although i have to admit for this prod it was "maybe" worth it! ;) also you had your 64k-tool handy anyway, so i guess you did fine! :p
gratz on 1st place btw!
the rendermonkey-project is obsolete since 2007 or sth...so contacting them will yield in no help i guess ;)
just have an eye on the bottom line, it tells you about the ALU-count etc.
so you just have to know about what instruction takes what amount of ALU-cycles and you should be fine! (traditional democoding! YAY!)
my problem with rendermonkey is the exeeding of registers...its a mistery to me how they solved that, if a shader is working in RM it is possible it is NOT working in your c++-counterpart sometimes, due to the ShaderCompilers seem to be different between RM and DirectX! this disqualified the most-best scenes i ever had and i still hate having gone the route of obsolete RenderMonkey!
in other words: dont trust in obsolete 3rdParty-Software!
also: making a tool for 4k is overEngineering, although i have to admit for this prod it was "maybe" worth it! ;) also you had your 64k-tool handy anyway, so i guess you did fine! :p
gratz on 1st place btw!
A sshot of a tool now ditched :)
Ahh... that felt good. Now it's out of sight and out of mind :)
maytz: damn... it looks nice! Too bad it was discontinued
maytz/lolnie: you should have releaed it as was :p would have thumbed down! for graphics, maybe sound would have made a change!
abductee: claim back to d3dx_37.dll and be fine, old shaderCompiler.dll there ;)
abductee: claim back to d3dx_37.dll and be fine, old shaderCompiler.dll there ;)
My TUM2012 "fast entry", done between two compos :)
maytz: that's just awesome. Gorgeous stuff.
Custom assembler for effects... with 256 variables space, limitation just4fun!
GP2X & asm rulez! :)
GP2X & asm rulez! :)
@hardy: thanks man!
unfortunately the 64k tool and the 4k tool are completly differnt ballgames, so not much help there :/
about the d3dx_37.dll, well i don't know man, from a personal standpoint i never really liked dx, but that is only me.
unfortunately the 64k tool and the 4k tool are completly differnt ballgames, so not much help there :/
about the d3dx_37.dll, well i don't know man, from a personal standpoint i never really liked dx, but that is only me.
Nice details abductee. I really wonder why RenderMonkey acts that shitty when hitting the instruction limit. I was kind a disappointed as you told us that yesterday at the SCSM.
^^
poo smear effect?
jesus, so the poop is still being smeared! gross!
So ein Lötzinn!
Custom profiler
u2: nice ui! I always envied people the ability to design nifty-looking user interfaces :)
OK:
Nice!
Actual work in progress on intro I am writing.
