|
His own greed made him run away from his friends and colleagues in order to keep all the money for himself. He's literally the old miser that fairy tales are written about.
|
#
?
Aug 13, 2018 07:01
|
|
|
|
| # ? May 18, 2024 02:11 |
|
|
believe me it's not common i feel sorry for notch  just when i'm feeling poorly or bitter about mental health care, I guess. he can just gently caress off otherwise especially considering his problems are self-inflictedas for exapunks, just got to the weird multiplayer thing. not gonna participate much but that's a neat idea
|
|
#
?
Aug 13, 2018 07:05
|
|
|
This game would be so much easier if the workers could aquire locks on m, I swear half of what I'm doing at the moment is carefully timing communication of subprocesses. You can't even do "workers wait to receive a packet from the master, then the one that receives it gets to send" because when that worker tries to send the packets will get picked up by the other workers waiting to receive the go signal. Local mode doesn't solve this because there's no way to test which other exas are in the same host as you so instead it's just all about very very careful timing. My baseball stats solution has 20+ noops in (the exact number obtained by trial and error) solely for the purpose of the workers not contenting for m gonadic io fucked around with this message at 09:10 on Aug 13, 2018 |
|
#
?
Aug 13, 2018 09:04
|
|
|
the loving tape drives kill me. made the nastiest solution. this is reminding me of TIS where I'm always thinking how good it would be to have just one more register. such a cool game though and the setting is awesome
|
|
#
?
Aug 14, 2018 00:32
|
|
|
i think i thought my solution was good for that one just did bank one. am super pleased with self coz initially i thought it was impossible but then later on came back and solution seemed obvious 😎
|
|
#
?
Aug 14, 2018 02:48
|
|
|
I'm fine with the register number but that shared memory bus with no locking and target header is truly something awful also getting hosed with baseball order of operations. Just tell me the number you want me to haaaave so I can reverse engineer that stupid math
|
|
#
?
Aug 14, 2018 02:50
|
|
|
gonadic io posted:This game would be so much easier if the workers could aquire locks on m, I swear half of what I'm doing at the moment is carefully timing communication of subprocesses. Unless you really really need to have these things in different nodes, you can use local mode just fine. Master sends a global message then listens for a local response, workers listen for global messages and send local responses. Even if you do need them in different nodes, you can sometimes have workers run home to send back their local response. Or even do things purely locally, by having both master and worker move to a different node for getting the response. (Admittedly, yes, that murders your activity score.)
|
|
#
?
Aug 14, 2018 03:00
|
|
|
echinopsis posted:i think i thought my solution was good for that one just went back and crushed it. thought of the right way to do it while i was poopin at work
|
|
#
?
Aug 14, 2018 07:52
|
|
|
jabor still got 10 cycles on me though 
|
|
#
?
Aug 14, 2018 07:52
|
|
|
i love -funroll-looping all my solutions
|
|
#
?
Aug 14, 2018 11:38
|
|
|
This game is extraordinarily difficult. Super fun tho
|
|
#
?
Aug 19, 2018 12:21
|
|
|
where r u up 2? im doing the redshift I think ive found the solution conceptually now its just a matter of fiddling with the odd line to make it actually work lol tho easy to find the code once I worked out what swiz meant
|
|
#
?
Aug 19, 2018 20:17
|
|
|
not far at all  I also can’t get the zine to print right, all the inverse sides are upside down 
|
|
#
?
Aug 19, 2018 20:57
|
|
|
what i love about this game and you’ll find the same is that puzzles seem impossible at first but you manage to progress thru thought until you can achieve them. it’s satisfying and you’ll get there believe me if i can do it you can
|
|
#
?
Aug 19, 2018 23:39
|
|
|
echinopsis posted:what i love about this game and you’ll find the same is that puzzles seem impossible at first but you manage to progress thru thought until you can achieve them. it’s satisfying and you’ll get there yeah every level i start by thinking "gently caress this is going to be a pain" but i always finish them feeling clever. im really surprised how small but versatile the instruction set is
|
|
#
?
Aug 19, 2018 23:45
|
|
|
what I've found I have a general idea of how to solve the puzzle but the nuances of synchronization and memory access aren't apparent until I get into it. I've also noticed that my solutions tend to be pretty short code wise but can take monstrous amounts of time
|
|
#
?
Aug 20, 2018 00:09
|
|
|
Just finished the Berkeley level and I feel proud of myself even though I’m like last in the rankings. this is a better game than Shenzhen IO even though it is similar, instruction limits were just so drat annoying. in this I feel like I’m writing state machines for the whole puzzle which is super fun.
|
|
#
?
Aug 20, 2018 05:34
|
|
|
I can’t believe how hard my computer has to work to run it though. It’s so simple, why can’t he write games that aren’t insanely cpu intensive.
|
|
#
?
Aug 20, 2018 05:35
|
|
|
Stereotype posted:I can’t believe how hard my computer has to work to run it though. It’s so simple, why can’t he write games that aren’t insanely cpu intensive. self-hosting the compiler took up all the time zach could have spent on optimising the cpu usage
|
|
#
?
Aug 20, 2018 08:33
|
|
|
Stereotype posted:I can’t believe how hard my computer has to work to run it though. It’s so simple, why can’t he write games that aren’t insanely cpu intensive. works on my machine   also i think it all boils down to javascript so you can imagine the hell that would look like
|
|
#
?
Aug 20, 2018 12:29
|
|
|
Stereotype posted:Just finished the Berkeley level and I feel proud of myself even though I’m like last in the rankings. well there are line limits, they are just really big. I hit the 50 line limit on the motorway sign level because I miscounted the number of letters and kept writing even wronger code trying to fix this. The line limit at least prompted me to go back and start again tho lol.
|
|
#
?
Aug 20, 2018 17:27
|
|
|
*gets back from week long camping trip* *does two tutorials* *quits game and turns on summerslam*
|
|
#
?
Aug 20, 2018 17:33
|
|
|
my bitter bi rival posted:*gets back from week long camping trip* lol yeah I bought the game but I can't bring myself to play it on the weekends; i'm so burned out at my job I can't focus on anything more complicated than battlefield, the world's dumbest game. I had an onsite last week though that went reasonably well so if I get that job I'm gonna take a week off to reset and probably play this a bit, I just need to get out of this funk
|
|
#
?
Aug 20, 2018 17:38
|
|
|
don't tell my employer but you can fairly accurately graph my mental health based on the commit history of my hobby poo poo.
|
|
#
?
Aug 20, 2018 17:40
|
|
|
Also I kind of wish you could interact with words more in this game - swiz them to parse numbers, use inline constants instead of being forced to read from a file at start even if it never changes etc
|
|
#
?
Aug 20, 2018 17:42
|
|
|
my bitter bi rival posted:*gets back from week long camping trip* ADINSX posted:lol yeah I bought the game but I can't bring myself to play it on the weekends; i'm so burned out at my job I can't focus on anything more complicated than battlefield, the world's dumbest game. i like this game cos my job is boring enterprise poo poo and this is like fun assembly janitoring. whomst among us doesnt yearn for the time when that xor trick to swap two variables around in c was actually relevant
|
|
#
?
Aug 21, 2018 05:02
|
|
|
okay just did the atm one and figured it out pretty quick
|
|
#
?
Aug 21, 2018 07:02
|
|
|
Stereotype posted:okay just did the atm one and figured it out pretty quick Did you just loop by 20, I keep meaning to go back to that one and unroll the loop until you hit the instructions limit. Otherwise it takes like a solid 5 mins to run the 100 tests
|
|
#
?
Aug 21, 2018 08:23
|
|
|
gonadic io posted:Did you just loop by 20, I keep meaning to go back to that one and unroll the loop until you hit the instructions limit. Otherwise it takes like a solid 5 mins to run the 100 tests Maybe we're talking about different ones, but I just loaded the #CASH register into X and then looped copying 20 to #DISP until X=0. It takes awhile because there is lots of cash in those things I guess. Also you make child EXAs to go do it so they're all done in parallel. Is there a later ATM one?
|
|
#
?
Aug 21, 2018 20:10
|
|
|
Stereotype posted:Maybe we're talking about different ones, but I just loaded the #CASH register into X and then looped copying 20 to #DISP until X=0. It takes awhile because there is lots of cash in those things I guess. we're talking about the same one. when i say loop unrolling i mean instead of being like is it 20 if yes dispense, loop if no finish i.e. you test the register once per $20 you can do is it 200 if yes then dispense, dispense, dispense, dispense, dispense, dispense, dispense, dispense, dispense, dispense then loop if no is it 20 if yes then dispense then loop if no finish then boom you've done 2 tests for $200 i.e. 0.2 tests per $20
|
|
#
?
Aug 21, 2018 20:26
|
|
|
gonadic io posted:we're talking about the same one. Ohhhh okay I get it, that's a lot more clever. That would be pretty trivial with 'generate' statements like in VHDL, then you just do like powers of two from the max (probably 512) downto zero? Actually you could probably create a slave EXA that only toggles the dispense key and doesn't ever do tests, then a master EXA that keeps track of cycles and kills it when it gets close? Stereotype fucked around with this message at 20:36 on Aug 21, 2018 |
|
#
?
Aug 21, 2018 20:34
|
|
|
gonadic io posted:Did you just loop by 20, I keep meaning to go back to that one and unroll the loop until you hit the instructions limit. Otherwise it takes like a solid 5 mins to run the 100 tests  it didnt take more than 30 seconds but I believe thats just processor speed maybe lol I have a huge amount of cycles compared to anyone else tho
|
|
#
?
Aug 21, 2018 20:41
|
|
|
I did what I always do with these games which is come out too strong out of the gate then get burned out and not want to turn my desktop on for a while.
|
|
#
?
Aug 21, 2018 22:07
|
|
|
Stereotype posted:Ohhhh okay I get it, that's a lot more clever. That would be pretty trivial with 'generate' statements like in VHDL, then you just do like powers of two from the max (probably 512) downto zero? do a few more levels and you'll get a tool to make stuff like that easier
|
|
#
?
Aug 21, 2018 22:19
|
|
|
is there a repository of people’s solutions? i have virtually no idea of how to optimise, until i see someone else’s solutions and then what they did seems obvious
|
|
#
?
Aug 21, 2018 23:24
|
|
|
echinopsis posted:
don't do a loop with subi then a test on x, just do a subi on T and loop a tjmp which will fall through when it's 0
|
|
#
?
Aug 21, 2018 23:28
|
|
|
gonadic io posted:Also I kind of wish you could interact with words more in this game - swiz them to parse numbers, use inline constants instead of being forced to read from a file at start even if it never changes etc can someone break down swizzing for a dumby like me
|
|
#
?
Aug 22, 2018 00:06
|
|
|
it lets you remap or mask the different digits in a number. so like you could divide by 10 by moving digit 4 to 3, 3 to 2, 2 to 1, and 1 to zip. likewise you could multiply by 10, or shove 4 small values in to on large number then read an individual value back out, reverse the digits, etc.
|
|
#
?
Aug 22, 2018 00:12
|
|
|
Bhodi posted:don't do a loop with subi then a test on x, just do a subi on T and loop a tjmp which will fall through when it's 0 oh yeah someone said this trick a few pages ago and I forgot about it.
|
|
#
?
Aug 22, 2018 01:54
|
|
|
|
| # ? May 18, 2024 02:11 |
|
|
Fuzzy Mammal posted:it lets you remap or mask the different digits in a number. so like you could divide by 10 by moving digit 4 to 3, 3 to 2, 2 to 1, and 1 to zip. likewise you could multiply by 10, or shove 4 small values in to on large number then read an individual value back out, reverse the digits, etc. I was misunderstanding the operation but I think i see it now - the example said 6789 and a mask of 3333 gives you "7777", which made no sense to me until I realized the order of digits is 4321, not 1234. i get it now!!
|
|
#
?
Aug 22, 2018 02:14
|
|