|
myhf, your 101 on the Animated Esports Sign impresses me. I am at 126. I suspect there's a nuance to power readings that I'm missing, unless I'm not seeing something in the data. Though, to get 126 I'm pretty sure you already have to have figured out that one of the three drink lines can be created via logic gates with the rest, and the click lines can invert each other.. Each instruction per chip executed costs 1 power, with the exception being that gen takes double what a mov slp mov slp cycle would?
|
#
?
Oct 19, 2016 07:05
|
|
|
|
| # ? Jun 10, 2024 15:00 |
|
|
nuru posted:myhf, your 101 on the Animated Esports Sign impresses me. I am at 126. I suspect there's a nuance to power readings that I'm missing, unless I'm not seeing something in the data. Though, to get 126 I'm pretty sure you already have to have figured out that one of the three drink lines can be created via logic gates with the rest, and the click lines can invert each other.. I got 126 without logic gates. e:Also, gen doesn't take any more power.
|
|
#
?
Oct 19, 2016 07:15
|
|
|
Dr. Stab posted:e:Also, gen doesn't take any more power. Hm, guess you're right! My earlier test must have had something else I missed.
|
|
#
?
Oct 19, 2016 07:21
|
|
|
gen is basically always equivalent to mov, slp, mov, slp. Notably, even if one of the values is 0, you still pay 1 power for the corresponding slp 0.
|
|
#
?
Oct 19, 2016 07:28
|
|
|
nuru posted:myhf, your 101 on the Animated Esports Sign impresses me. I am at 126. I suspect there's a nuance to power readings that I'm missing, unless I'm not seeing something in the data. Though, to get 126 I'm pretty sure you already have to have figured out that one of the three drink lines can be created via logic gates with the rest, and the click lines can invert each other.. There are 46 frames of output, so the necessary MOV and SLP operations take 92 power. The 10-frame cycle takes 20 operations, which doesn't align neatly with 14-line MC or ROM size, so you need to take measures to realign them or hand off control.
|
|
#
?
Oct 19, 2016 07:56
|
|
|
Phssthpok posted:There are 46 frames of output, so the necessary MOV and SLP operations take 92 power. The 10-frame cycle takes 20 operations, which doesn't align neatly with 14-line MC or ROM size, so you need to take measures to realign them or hand off control. This made me revisit the original puzzle and discover that one of my SA goon friends (Scaevolus) has a 49 for power on this puzzle. I'll be over here, being mystified.
|
|
#
?
Oct 19, 2016 08:04
|
|
|
Ignoranus posted:This made me revisit the original puzzle and discover that one of my SA goon friends (Scaevolus) has a 49 for power on this puzzle. I'll be over here, being mystified. I think he mentioned earlier in this thread that you can do the clicks with 0 power, using pure-logic oscillators. If you try running just the drinks with no clicks, you will get the same score. I'm not sure whether it still works with the recent changes. Now I've overhauled my feng shui to fit two MC6000s on the board, and gotten it down to 93 power. Using the program counter wraparound to get an arbitrary modulo operation is a good trick from TIS-100 graphics levels.
|
|
#
?
Oct 19, 2016 08:11
|
|
|
I really don't like the way you can make an oscillator out of logic gates. It's probably intended, but it just feels creepy and weird like a perpetual motion machine. It feels like with the way everything else in the game works that it should cause a "waiting for sleep" error the same way the IC's do. It also seems like maybe logic gate state changes should consume power. Being able to slap a not gate down and instantly get a 100 over Simple I/O while consuming no power feels wrong. Either Zachtronics should remove the logic gates altogether or make them cost power so that there's no 0 power solutions possible. I guess it's fundamentally space-limited and if you want to do that then you're pretty much giving up on the cost metric. I guess that kind of balances it. I just would like to see at least the not gate use some power.
|
|
#
?
Oct 19, 2016 09:49
|
|
|
ErIog posted:I really don't like the way you can make an oscillator out of logic gates. It's probably intended Phssthpok posted:I think he mentioned earlier in this thread that you can do the clicks with 0 power, using pure-logic oscillators. If you try running just the drinks with no clicks, you will get the same score. I'm not sure whether it still works with the recent changes.
|
|
#
?
Oct 19, 2016 11:07
|
|
|
Bonfire Lit posted:I'm not sure it is, considering that sending XBus messages or sometimes just using a completely separate MC breaks it. I agree with you that it doesn't 100% seem intended considering it works exactly the opposite of the way the game says that signals operate, but I didn't want to be presumptuous. I anticipated someone replying to me saying, "tough poo poo, git gud newb." Basically, in every other respect the game treats signal propagation as instantaneous, but the way the discrete logic components work together with each other causes that not to be 100% true. I just fundamentally think that with the entire rest of the game taken as context, that signals should not be allowed to oscillate like that across time steps without the input changing. If you understand the way the game is supposed to work you would never in a million years think that kind of thing would be allowed. So that's why I have such a distaste for it. It feels like unless you play 200 hours of it and go hunting for weirdness -or- see a spoiler solution here or on Reddit that you would never naturally come up with that set of gates. In the real world I'm sure there's funky poo poo with signal propagation that allows you to make hosed up timers like that, but it seems like this game goes out of it's way to take that out of the equation in order to make a good video game except in this instance where you can slop together a bunch of them on a board and get a perpetual oscillator for 0 power. ErIog fucked around with this message at 11:23 on Oct 19, 2016 |
|
#
?
Oct 19, 2016 11:18
|
|
|
It's sounding like the thing I thought about trying to build with gates (fets / flip flops) but assumed they would never work due to how simple io is designed nay actually work after all?
|
|
#
?
Oct 19, 2016 14:31
|
|
|
nuru posted:It's sounding like the thing I thought about trying to build with gates (fets / flip flops) but assumed they would never work due to how simple io is designed nay actually work after all? Yeah I think you can genuinely build a latch. It'll take up a lot of space. I genuinely can't say I like it either - you should have to converge your signals before advancing a time tick, or maybe they should only propagate once per time tick.
|
|
#
?
Oct 19, 2016 16:09
|
|
|
Speaking of which, is there eventually a sandbox that lets you profile power numbers?
|
|
#
?
Oct 19, 2016 18:38
|
|
|
Man, I don't know why but the Aquaponics Robot puzzle is somehow mystifying to me (as was Cool Dad, honestly). Handling the non-blocking input is aggravating because I have to test the incoming values non-destructively so it seems like the only real way is to move the value on the xBus input into ACC each tick and THEN test it, because if I just test the incoming value it's destroyed and I have effectively lost part of the packet. Does this seem right or am I missing some sort of cleverness regarding packet handling?
|
|
#
?
Oct 20, 2016 01:44
|
|
|
If there's a better way, I don't know about it.  If your crunching unit would thrive on not having to spend precious lines polling everything that comes down the XBus tube, you could have a spare MC4000 just to pluck out the packets from the -999 static and keep the actual doing-things chip sleeping until something comes up.(gently caress efficiency)
|
|
#
?
Oct 20, 2016 02:05
|
|
|
I AM BECOME IMMORTAL
|
|
#
?
Oct 20, 2016 10:22
|
|
|
A couple of little updates:quote:Update: Quick fix for the hang on exit (Mac and Linux)
|
|
#
?
Oct 21, 2016 01:53
|
|
|
Hah, the limited edition is already out-of-date. I mean it was inevitable, but it's still funny.
|
|
#
?
Oct 21, 2016 02:28
|
|
|
The TRM being out of spec officially makes the game my day job.
|
|
#
?
Oct 21, 2016 02:36
|
|
|
Jabor posted:Hah, the limited edition is already out-of-date. Documentation changing or not being accurate or requiring you to make handwritten notes is pretty realistic, imo. It's a binder. Just print your own changes and shove them in. The more poo poo you have to print to keep it up to date the more authentic the Limited Edition will feel. 
ErIog fucked around with this message at 02:40 on Oct 21, 2016 |
|
#
?
Oct 21, 2016 02:38
|
|
|
They should make errata sheets to print off and stick in the binder.
|
|
#
?
Oct 21, 2016 02:40
|
|
|
Type your own, making sure to use mismatched text formatting for verisimilitude.
|
|
#
?
Oct 21, 2016 03:19
|
|
|
http://imgur.com/a/1QwVX Anyone see a reason why this shouldn't work? Seems like a bug in the interaction between non-blocking XBus and the slx instruction. The "slx x0" instruction never triggers on other IC. I tried to e-mail Zach about it as a bug report, but it seemed like he misunderstood and told me to go ask Steam Forums or Reddit to debug my code. So.. can anyone tell me why this doesn't work even though it totally seems like it should? Zach from Zachtronics posted:You should post on the Steam forums or Reddit about this… they’ll be able to help you debug it and figure out what is going wrong. Also, lol, he assumes I don't know how to rotate components. Is my feng shui that bad?  GotLag posted:Type your own, making sure to use mismatched text formatting for verisimilitude. This is actually one of the only things that disappointed me about the Limited Edition. There's too much coherence in terms of fonts between what are supposed to be disparate pieces of documentation. ErIog fucked around with this message at 04:35 on Oct 21, 2016 |
|
#
?
Oct 21, 2016 04:29
|
|
|
ErIog posted:http://imgur.com/a/1QwVX
|
|
#
?
Oct 21, 2016 04:45
|
|
|
Thanks for the sanity check. That's basically what I said in my reply to his e-mail. If that's intended then that's fine I guess, but it felt like an oversight which is why I called it a "bug."
|
|
#
?
Oct 21, 2016 04:47
|
|
|
I feel retarded. I can't even come up with a workable concept for the control router. Having to transmit the destination and size fields as well is killing me.
|
|
#
?
Oct 21, 2016 06:06
|
|
|
GotLag posted:I feel retarded. I can't even come up with a workable concept for the control router. Having to transmit the destination and size fields as well is killing me. If you have a dedicated MC for each output, you can connect all of them to the main input bus and only have to count the input length once. Consider this design with a "watchdog" looking for the first packet, and a "multiplexer" that wakes the correct output MC. Then each output MC just has to pass the ID through (2 lines), store the length and output it (2 lines), loop and output each byte (4 lines). I like this design because the feng shui is clean and every output MC can have the exact same code. But you need DAT registers and way more instruction space to really optimize this level. 
Phssthpok fucked around with this message at 07:44 on Oct 21, 2016 |
|
#
?
Oct 21, 2016 07:22
|
|
|
i was beating my head against the bitcoin terminal until I muddled it through. the automatic window was a refreshing change that made think "me am smart" so that was nice 
|
|
#
?
Oct 21, 2016 08:24
|
|
|
Phssthpok posted:If you have a dedicated MC for each output, you can connect all of them to the main input bus and only have to count the input length once. Consider this design with a "watchdog" looking for the first packet, and a "multiplexer" that wakes the correct output MC. Then each output MC just has to pass the ID through (2 lines), store the length and output it (2 lines), loop and output each byte (4 lines). Here's a tip about one of those IC's. You can axe one of them while keeping the same concept.
|
|
#
?
Oct 21, 2016 08:33
|
|
|
Phssthpok posted:If you have a dedicated MC for each output, you can connect all of them to the main input bus and only have to count the input length once. I'll just blame being at home sick for not being able to solve this. In return I will provide a tip for Replacement Factory Module: one of the possible inputs can be solved for no power cost GotLag fucked around with this message at 08:45 on Oct 21, 2016 |
|
#
?
Oct 21, 2016 08:42
|
|
|
This game is really good at making me feel like an idiot. Here's my first solution for Remote Kill Switch:  Here's my solution after I realized "dst" would actually come in useful for once:  Still much slower than the goon solutions, but good enough for me!
|
|
#
?
Oct 21, 2016 09:57
|
|
|
How did someone do the control router in ¥5? It seems like there would be nowhere to put the key.
|
|
#
?
Oct 21, 2016 10:22
|
|
|
Reading the security chip via one of the output pins? Edit: hosed if I know how they fit the code in GotLag fucked around with this message at 11:25 on Oct 21, 2016 |
|
#
?
Oct 21, 2016 11:11
|
|
|
GotLag posted:In return I will provide a tip for Replacement Factory Module: one of the possible inputs can be solved for no power cost
|
|
#
?
Oct 21, 2016 19:43
|
|
|
Zach is taking puzzle submissions for a second/bonus campaign, but the exact details are a spoiler for those that haven't finished the game. In either case I thought I'd link the details here in case there are people here that haven't seen it.
|
|
#
?
Oct 21, 2016 20:17
|
|
|
MATHS Can you spot the desperate botch to save my "divide by 10" block?
|
|
#
?
Oct 21, 2016 21:51
|
|
|
Edit: Nevermind, I am dumb and can't count. Got it.
WhiteHowler fucked around with this message at 22:15 on Oct 21, 2016 |
|
#
?
Oct 21, 2016 22:03
|
|
|
homeless snail posted:Two, actually I wasn't counting the zero case.  
GotLag fucked around with this message at 01:52 on Oct 22, 2016 |
|
#
?
Oct 22, 2016 01:35
|
|
|
Kaislioc posted:Zach is taking puzzle submissions for a second/bonus campaign, but the exact details are a spoiler for those that haven't finished the game. In either case I thought I'd link the details here in case there are people here that haven't seen it. Ooh, I have Ideas. Should be interesting.
|
|
#
?
Oct 22, 2016 03:56
|
|
|
|
| # ? Jun 10, 2024 15:00 |
|
|
I managed to stack all numbers before stacking any dragons. This would make a good achievement.
|
|
#
?
Oct 22, 2016 08:03
|
|