Members in Shoutbox
None.

Shoutbox Search
Search for:


Shoutbox Commands
/w [name] > Whisper
/r > Reply to last whisper
/me > Marks as action

Shoutbox Information
Moderators may delete any and all shouts at will.
Global Shoutbox
Please log in to shout.
Pages: < 1 « 1190 1191 1192 1193 11943571 >

[2016-12-05. : 3:28 am]
jjf28 -- CecilSunkure
CecilSunkure shouted: If you can't feed that bytecode as-is to the hardware, then you require a compiler
in the case of direct interpretation there is no process of creating machine code that did not previously exist, to be fed to the hardware. The entire JVM is written, no new code is generated, the register values are different and existing routines are invoked in new and interesting ways
[2016-12-05. : 3:27 am]
CecilSunkure -- name your stacks. I'll give u y/n
[2016-12-05. : 3:27 am]
lil-Inferno -- ok
[2016-12-05. : 3:27 am]
CecilSunkure -- lil-Inferno
lil-Inferno shouted: CecilSunkure w/e, getting fatter $tacc$ than you so who cares
doubt it.
[2016-12-05. : 3:27 am]
lil-Inferno -- except you C|X^ D
[2016-12-05. : 3:27 am]
lil-Inferno -- CecilSunkure
CecilSunkure shouted: so i guess if u just write basic bitch code it doesn't matter if u use java
w/e, getting fatter $tacc$ than you so who cares
[2016-12-05. : 3:27 am]
CecilSunkure -- no u
[2016-12-05. : 3:26 am]
Moose -- Cecil, how much catv you read and how much Plan9 you running on your machine LMAO
[2016-12-05. : 3:22 am]
lil-Inferno -- ok
[2016-12-05. : 3:22 am]
Moose -- Probs don't even know the GNU toolchain SMFH
[2016-12-05. : 3:22 am]
Dem0n -- w/e
[2016-12-05. : 3:22 am]
CecilSunkure -- They prance around believing in fantasies of free tradeoffs and infinite machine power
[2016-12-05. : 3:22 am]
CecilSunkure -- SMFH this is why I hate OOP programmers
[2016-12-05. : 3:21 am]
Moose -- new memes compiling as we speak O_O
[2016-12-05. : 3:21 am]
CecilSunkure -- Well you didn't admit you were wrong, so I'm still on the topic of you didn't know what a compiler was
[2016-12-05. : 3:20 am]
jjf28 -- oh I'm on a slightly different topic
[2016-12-05. : 3:20 am]
CecilSunkure -- Not in the context of defining a compiler, and not being free
[2016-12-05. : 3:20 am]
jjf28 -- CecilSunkure
CecilSunkure shouted: Doesn't matter how tiny or clever those performance hits are
uhh, yeah, it does xD
[2016-12-05. : 3:19 am]
CecilSunkure -- It's still a compiler and it's not free
[2016-12-05. : 3:18 am]
CecilSunkure -- Doesn't matter how tiny or clever those performance hits are
[2016-12-05. : 3:18 am]
jjf28 -- the first time any given method runs; it's really not a big deal, memory management gets pretty icky tho
[2016-12-05. : 3:18 am]
CecilSunkure -- run-time compiler*
[2016-12-05. : 3:17 am]
CecilSunkure -- Therefor, any language that uses a compiler will take a performance hit.
[2016-12-05. : 3:17 am]
CecilSunkure -- And compilers are not ever free, by definition.
[2016-12-05. : 3:16 am]
CecilSunkure -- If you can't feed that bytecode as-is to the hardware, then you require a compiler
[2016-12-05. : 3:16 am]
jjf28 -- it's not turning the code into statements used by the processor, the statements the processor uses are already written
[2016-12-05. : 3:16 am]
CecilSunkure -- When we read bytecode, what happens? Eventually, DOESN'T MATTER HOW, it is translated into machine code, because modern processors require machine code to do shit
[2016-12-05. : 3:15 am]
CecilSunkure -- "A compiler is a special program that processes statements written in a particular programming language and turns them into machine language or "code" that a computer's processor uses."
[2016-12-05. : 3:14 am]
CecilSunkure -- We talked about that already in the SB
[2016-12-05. : 3:14 am]
jjf28 -- they're not compiling the assembly into machine code that runs on the processor, they're running the code on the simulation program
[2016-12-05. : 3:14 am]
CecilSunkure -- *ahem* you said simulator
[2016-12-05. : 3:13 am]
jjf28 -- haven't you seen assembly simulators?
[2016-12-05. : 3:13 am]
Moose -- back to dictionary debates SMDH
[2016-12-05. : 3:13 am]
CecilSunkure -- ok
[2016-12-05. : 3:13 am]
jjf28 -- no it's not
[2016-12-05. : 3:13 am]
CecilSunkure -- That layer of abstraction is called a compiler
[2016-12-05. : 3:13 am]
jjf28 -- CecilSunkure
CecilSunkure shouted: Interpreter running bytecode is still interpretation. It's a compiler.
no, it's not turning it into machine code, it's processing the bytecode manually -- a whole layer of abstraction
[2016-12-05. : 3:12 am]
CecilSunkure -- The only way it wouldn't be a compiler is if there's no performance hit because it's just machine code
[2016-12-05. : 3:12 am]
Moose -- jjf28
jjf28 shouted: CecilSunkure they're real
your FACE is real
[2016-12-05. : 3:12 am]
CecilSunkure -- Interpreter running bytecode is still interpretation. It's a compiler.
[2016-12-05. : 3:12 am]
jjf28 -- CecilSunkure
CecilSunkure shouted: That's 1 case, and quite a fantasy at that
they're real
[2016-12-05. : 3:11 am]
jjf28 -- they do have those, also the interpreter can straight up run the bytecode
[2016-12-05. : 3:11 am]
CecilSunkure -- That's 1 case, and quite a fantasy at that
[2016-12-05. : 3:11 am]
CecilSunkure -- what a hardware bytecode processor? LMAO
[2016-12-05. : 3:11 am]
CecilSunkure -- hire*
[2016-12-05. : 3:10 am]
jjf28 -- CecilSunkure
CecilSunkure shouted: jjf28 Bytecode still needs to be compiled. All bytecode does is remove heftier front-end parsing. It's still interpreted. JIT is pretty cool, but at that point really why not use C? Interpretation for cross platform is silly, when interpretation can be much more powerful for fast iteration time... Cross-platform at the cost of performance? Really? I'd argue a good C programmer can shit out features faster than any Java goon, and have it still be cross-platform, without any performance hits.
I just told you the cases where it doesn't need to be compiled :lol:
[2016-12-05. : 3:09 am]
CecilSunkure -- Mythical man month has an answer for that
[2016-12-05. : 3:09 am]
CecilSunkure -- It's like where is the tradeoff? Should we just require good C programmers? Or should we take a performance hit and higher lesser goons...
[2016-12-05. : 3:08 am]
CecilSunkure -- jjf28
jjf28 shouted: CecilSunkure jjf28 CecilSunkure an interpreter could just run the byte code itself, or the hardware could have a built in byte-code processor, or ofc, JIT compilation could be used - the exact behavior is platform dependent but even if it's JIT compilation there's a big difference from C-style compilation
Bytecode still needs to be compiled. All bytecode does is remove heftier front-end parsing. It's still interpreted. JIT is pretty cool, but at that point really why not use C? Interpretation for cross platform is silly, when interpretation can be much more powerful for fast iteration time... Cross-platform at the cost of performance? Really? I'd argue a good C programmer can shit out features faster than any Java goon, and have it still be cross-platform, without any performance hits.
[2016-12-05. : 3:06 am]
Moose -- Because embedded systems. Like, nobody putting the JVM on your microwave. Yet. X^D
[2016-12-05. : 3:05 am]
jjf28 -- CecilSunkure
CecilSunkure shouted: Compiling = interpreting a JAR file. fyi
jjf28
jjf28 shouted: not really
CecilSunkure
CecilSunkure shouted: What do you mean not really. That's exactly what interpreting is -- runtime compilation
an interpreter could just run the byte code itself, or the hardware could have a built in byte-code processor, or ofc, JIT compilation could be used - the exact behavior is platform dependent but even if it's JIT compilation there's a big difference from C-style compilation
[2016-12-05. : 3:05 am]
Dem0n -- I guess literally all android phones are unusable pieces of shit X^ D
[2016-12-05. : 3:04 am]
CecilSunkure -- so i guess if u just write basic bitch code it doesn't matter if u use java
[2016-12-05. : 3:04 am]
CecilSunkure -- like FFT code
[2016-12-05. : 3:03 am]
CecilSunkure -- Dem0n
Dem0n shouted: CecilSunkure why do people act like we still use computers with like 64 mbs ram and shit. like this matters LMAO
It only does for certain pieces of code
[2016-12-05. : 3:03 am]
Dem0n -- CecilSunkure
CecilSunkure shouted: So Java: compile your program every time it's run!
why do people act like we still use computers with like 64 mbs ram and shit. like this matters LMAO
[2016-12-05. : 3:03 am]
CecilSunkure -- I'm now more interested in watching this guy write than explain FT LMAO
[2016-12-05. : 3:02 am]
Moose -- Finally jogged my memory to remember where it was LMAO
[2016-12-05. : 3:01 am]
CecilSunkure -- cool thx!
[2016-12-05. : 3:01 am]
Moose -- https://www.nayuki.io/page/free-small-fft-in-multiple-languages oh ya, this was part of what I ended up using
[2016-12-05. : 3:00 am]
CecilSunkure -- https://www.youtube.com/watch?v=1JnayXHhjlg -- beautiful AF hand writing
[2016-12-05. : 3:00 am]
CecilSunkure -- interesting
[2016-12-05. : 2:59 am]
Moose -- Ya, he good AF tbh
[2016-12-05. : 2:58 am]
CecilSunkure -- bluestein
[2016-12-05. : 2:58 am]
Moose -- https://en.wikipedia.org/wiki/Fast_Fourier_transform ya, the ones from here
[2016-12-05. : 2:58 am]
CecilSunkure -- Are these algorithms converging a set of sin/cos waves?
[2016-12-05. : 2:57 am]
CecilSunkure -- So there's multiple flavors of algorithms to apply FFT? Cooley-Tukey made one, and you mentioned another?
[2016-12-05. : 2:56 am]
Moose -- The DFT isn't too bad to understand the computation of, but for the amount of depth I went into understanding, anything more advanced than the Cooley-Tukey FFT algorithm does magical bullshit that just works LMAO
[2016-12-05. : 2:54 am]
CecilSunkure -- I like it
[2016-12-05. : 2:53 am]
Moose -- FFT approximates a function (the signal) as a sum of sin/cos at different frequencies

Pages: < 1 « 1190 1191 1192 1193 11943571 >


Members Online: Moose, O)FaRTy1billion[MM]