You know, I looked at the Steam forum page for the tournament, months after I had to bow out, and it was still going on. It actually went until November. The people who play this game are seriously hardcore about it.
Anyway, I'm just posting to let people know that Infinifactory, Zachtronics' latest game, just hit Steam early access. From what I've played so far, it definitely has that Spacechem feel. I'm digging it. Despite it being early access, the main campaign is done. It's missing the "share solution" feature and whatever additional mechanics/levels they plan on adding later, but it's complete enough to play and enjoy.
Anyway, I'm just posting to let people know that Infinifactory, Zachtronics' latest game, just hit Steam early access. From what I've played so far, it definitely has that Spacechem feel. I'm digging it. Despite it being early access, the main campaign is done. It's missing the "share solution" feature and whatever additional mechanics/levels they plan on adding later, but it's complete enough to play and enjoy.
I'm 2 stages into it and am loving it. It's spacechem in 3d, and really quite a gem.
Anyway, I'm just posting to let people know that Infinifactory, Zachtronics' latest game, just hit Steam early access. From what I've played so far, it definitely has that Spacechem feel. I'm digging it. Despite it being early access, the main campaign is done. It's missing the "share solution" feature and whatever additional mechanics/levels they plan on adding later, but it's complete enough to play and enjoy.
I'm 2 stages into it and am loving it. It's spacechem in 3d, and really quite a gem.
That's how far in I got before I had to leave for work. Looking at the histogram and seeing folks completing it in half the cycles brought back that old feeling. From the quick look I got at puzzle 3, it seems the difficulty ramps up quickly.
Any takers on a new thread? I could make it when I get home, but I don't trust myself to keep it up to date. The game deserves a spotlight, as well as a chance to distance itself from Spacechem's chemistry theme. The devs think a good chunk of people were scared off, because of it. I'll do it, if I have to.
One thing I already miss from Spacechem is the ability work on solutions with just pencil and paper. The shift to 3d and more open layouts makes it harder to conceptualize solutions outside of the game. That said, I have an idea for a solution to the triple weld problem, assuming certain blocks work the way I think they do.
New tournament next month. Saw a post on the subreddit. Posting for those who don't follow it. I'm very tempted to participate, despite what it could do to my sleep schedule.
Fired up the game last night. It wasn't even installed, and I've had this pc since last summer. Managed to optimize a Researchnet puzzle(Exercise), before pulling myself away.
Well, congrats on that front. While not nearly in the same league as caring for a child, playing can get in the way if things I feel are important. On the puzzle from the other day, I stopped around 2 A.M., which isn't too far from when I'd normally call it, but it was late enough that I couldn't get up in time for my morning run.
The thing that makes me think I could participate and manage my time is remembering that I own the mobile version and a phone that can run it, so I can do most of the grunt work away from home.
Huh, good thing I still had this thread bookmarked. I don't seem to log in every day on the forums these days, so I could have easily missed this. Might need a reminder for this, but it'll be good to flex those SpaceChem muscles again.
Been seeing a lot of activity on the high scores thread, as well as on Youtube. Seems like a good number of people are warming up by optimizing the last batch of Researchnet puzzles. I worked on the Alcohol series and a few others. Looks like it's shaping up to be a fierce competition.
1st puzzle is up. Actually, the 1st and 0th puzzles. They're all going up on r/researchnet.
1 a week, with a little over a 2 weeks to complete each, which seems lenient. The tricky part is the scoring system. Since the overall score is your cycles v top cycles + your symbols v top symbols, and you can't know how well others did, your best bet is to make the solutions optimized for both categories, then make middle ground solutions and compare them to your extremes.
Thanks for the heads up. Apparently, there was a tournament and mini-event in 2015 and 2016, respectively? I feel like I'm gonna be way behind the curve in terms of proficiency from having not played for about two and a half years. Found my research notebook from 2014 and there's plenty of space for this year's event. Let's go!
Got puzzle 0 down to 1250/27. Didn't submit it, since it doesn't count, though it might've helped add some test data.
I did sumbit a solution for the 1st. It's about as solid as I can manage, without a major epiphany. Haven't seen any of the bragging that accompanied previous tournaments. Might be an effect of the scoring.
Edit: On 2nd thought, that doesn't make much sense. People knew the goals before. Knowing ballparks now might let you know if an alternate strategy was worth pursuing, but wouldn't make things any easier.
Also, I think I just had that epiphany.
Edit 2: Epiphany failed. What I had in mind would've put it on the slower track I started with. I have 1 other idea that might save some symbols.
Also, the results for puzzle 0 are out. Had I submitted, I would've tied for 6th. Of course, had it counted, more people would've submitted.
Edit 3: A different strategy has sped things up, but at too high of a symbol cost. Poor positioning is likely to blame. It's worth investigating further, at least.
So the 1st round is nearly over, the 2nd has been up for a week and the 3rd just dropped. How's everyone doing? I've submitted for both weeks. Trying to keep from working on more than 1 at a time, so I give myself a week and submit before the new puzzle arrives.
Man, I need to get around to the second puzzle this weekend. I already have some ideas of how I want to try out for it, though I don't know if I'll focus more on symbols or cycles. For the first puzzle, I focused more on cycles, and I got to a point where I felt seeking a better symbol-based solution would cost too many cycles to be worth it. The new scoring system is tough...
Yeah, it's tricky. You can calculate your own solutions, but it also makes sense to run the numbers against what you assume the best solutions will be. Did that with the 1st puzzle, then ended up beating my estimate for lowest cycles. Just barely, so it might've been a good guess, or I might've missed something.
The 3rd puzzle gives 1.5 extra points for getting 16 outputs with 15 inputs. Almost certainly worth it, since the max you could get otherwise is 2 points, which is already unlikely.
Hah, good thing you reminded me that the Steam community thread for the tournament exists, otherwise I would have forgotten that particle smashing was a thing, and a thing that is legal, at that. Too late for me to see if that's something that would benefit the 1st puzzle, but at least I can avoid targeting a 16 outputs from 15 inputs solution from the outset just to get something that works. I gotta review how to do particle smashing first, though.
11th out of 31 and the 3rd fastest solution. Not bad. My symbol count was atrocious. Looks like smashing and non-smashing were almost equally viable, taking 1st and 2nd place.
E: I just noticed I had both the worst symbol count and the worst smashing solution.
Ooh. I'm gonna need to check out Astronelson's top non-smash solution later. Meanwhile, I still haven't submitted a solution for puzzle 2 yet, since I'm having a hard time figuring out how to refine my initial standard solution. I might have to exploit sequence knowledge to get a better result.
I'm fairly happy with my solution for #2, which is standard. I figure the alternate solutions will be faster, but needing 12 outputs for each side will jack up the symbol count, unless there's an underlying pattern to exploit.
I could maybe get my solution to finish a bit faster, but I don't think the cycle savings would be worth the symbol cost.
E: Got it faster. Still don't think it's worth it.
E2: Worked out a solution for #3. Funny that the assumed standard method for solving this one is smashing, while using all of the inputs is the special case. I think Gggol gave up the game a bit by mentioning smashing in the post for that puzzle. Might not have occurred to try that if it wasn't.
MrBlarney and I stood side by side against the rise of the machines. We lost, but not before taking out a couple of those metal bastards. Surprised to see the top machine had the lowest symbols.
Got a solution for #4. The cascade of realizations about its hurdles was something else. I have a slightly improved version I could submit, but I want to try 1 other thing which might lead to greater gains.
I'm really hating the fact that I don't have much time to go beyond first-pass solutions this time around. I just couldn't invest enough time to map out a good precognition solution for Seven or Eight and then I got really worried about cycle symbol counts. But the fact that the best precog solution was also low symbols, and that there were some fairly short loopers is pretty amazing, yeah.
Yeah, I would've had to do it entirely on PC, since my phone has different rng. I think I'll take a crack at a 15 input solution for #3, now that I've put #4 to bed.
I don't know if #5 is even worth the effort, given the time it'll take to troubleshoot. If I do it at all, it'll be in pieces over the 2+ weeks, rather than within a week, like the previous puzzles.
I ended up working on #4 some more, instead of making a 15 input solution for #3. Even though I didn't like it at first, I like that it's the first puzzle here that doesn't have wildly branching solution paths. No smashing, state machines or special conditions. Just make the product and figure out what to do with the waste.
The herd is starting to thin. I'm listed as 16th for #4, but Gggol must be using an outdated list. The 642/86 solution was the 2nd one I sent. The last one was 634/83, which bumps my 1.002 to 1.026. Enough to squeak into 15th, barring improvements from other players.
#5 really is a slog. Not done yet, but I'm on the right track. It just takes so long, even at speed 4, to get to the point where you can work out the next step.
I had a pretty much last-minute solution for #4, 'cause I only worked on it on Sunday and had the longest time making a garbage-sorting method. On #3, I immediately saw where I went wrong on my solution and could have immediately jumped up to your level. Whoops.
As for #5, it doesn't seem too bad to me at first glance. You could just create a copy of the puzzle, substituting the Meitnerium for some other 6-bond atom like Iron, to be able to handle the construction part. Fusing upwards to Meitnerium though... yeah, that's where a lot of efficiency could be lost or won. Lots of flippity floppity math possible here.
Good idea about the substitute atom. I took my 109 counter and simplified it to 16, while keeping the structure there for space limitations. Once I finish the rest of it, I'll put the other flip-flops and syncs back into place. I'm not used to making such a large molecule, especially with only 2 bonders and no sensor. I get that it has to be done in chunks, but the issue is having 1 waldo doing the bulk of the work. Getting the other waldo in on the action is tricky, because you're dealing with 109, a large(for SC) prime. The counter I have has the benefit of using syncs, so I can get close and stick the extra onto the front/back. Can't do that on both waldos without significantly slowing things down. With a solution that's on track to end up around 500k cycles, that's saying something. I could maybe give the other waldo something simple to do every 8/16/32 cycles. Assuming it doesn't interfere with the workhorse waldo, it'd end up in sync.
E: Just realized what I initially had in mind is physically impossible. New track makes much more sense.
Posts
Anyway, I'm just posting to let people know that Infinifactory, Zachtronics' latest game, just hit Steam early access. From what I've played so far, it definitely has that Spacechem feel. I'm digging it. Despite it being early access, the main campaign is done. It's missing the "share solution" feature and whatever additional mechanics/levels they plan on adding later, but it's complete enough to play and enjoy.
I'm 2 stages into it and am loving it. It's spacechem in 3d, and really quite a gem.
Noooooooooooo damn you I have things to be doing why do you bring this to my attention
That's how far in I got before I had to leave for work. Looking at the histogram and seeing folks completing it in half the cycles brought back that old feeling. From the quick look I got at puzzle 3, it seems the difficulty ramps up quickly.
One thing I already miss from Spacechem is the ability work on solutions with just pencil and paper. The shift to 3d and more open layouts makes it harder to conceptualize solutions outside of the game. That said, I have an idea for a solution to the triple weld problem, assuming certain blocks work the way I think they do.
Edit: nvm
http://forums.penny-arcade.com/discussion/195938/infinifactory-play-spacechem-in-3d#latest
Fired up the game last night. It wasn't even installed, and I've had this pc since last summer. Managed to optimize a Researchnet puzzle(Exercise), before pulling myself away.
Unfortunately, I have a toddler now, and can't spare the time necessary.
Good luck to all who enter though!
The thing that makes me think I could participate and manage my time is remembering that I own the mobile version and a phone that can run it, so I can do most of the grunt work away from home.
Been seeing a lot of activity on the high scores thread, as well as on Youtube. Seems like a good number of people are warming up by optimizing the last batch of Researchnet puzzles. I worked on the Alcohol series and a few others. Looks like it's shaping up to be a fierce competition.
1 a week, with a little over a 2 weeks to complete each, which seems lenient. The tricky part is the scoring system. Since the overall score is your cycles v top cycles + your symbols v top symbols, and you can't know how well others did, your best bet is to make the solutions optimized for both categories, then make middle ground solutions and compare them to your extremes.
That's the 1st. If anyone's curious, I got 1251/39 on the 0th. Could've made it a bit faster, but it would've more than doubled the symbols.
I did sumbit a solution for the 1st. It's about as solid as I can manage, without a major epiphany. Haven't seen any of the bragging that accompanied previous tournaments. Might be an effect of the scoring.
Edit: On 2nd thought, that doesn't make much sense. People knew the goals before. Knowing ballparks now might let you know if an alternate strategy was worth pursuing, but wouldn't make things any easier.
Also, I think I just had that epiphany.
Edit 2: Epiphany failed. What I had in mind would've put it on the slower track I started with. I have 1 other idea that might save some symbols.
Also, the results for puzzle 0 are out. Had I submitted, I would've tied for 6th. Of course, had it counted, more people would've submitted.
Edit 3: A different strategy has sped things up, but at too high of a symbol cost. Poor positioning is likely to blame. It's worth investigating further, at least.
The 3rd puzzle gives 1.5 extra points for getting 16 outputs with 15 inputs. Almost certainly worth it, since the max you could get otherwise is 2 points, which is already unlikely.
E: I just noticed I had both the worst symbol count and the worst smashing solution.
I could maybe get my solution to finish a bit faster, but I don't think the cycle savings would be worth the symbol cost.
E: Got it faster. Still don't think it's worth it.
E2: Worked out a solution for #3. Funny that the assumed standard method for solving this one is smashing, while using all of the inputs is the special case. I think Gggol gave up the game a bit by mentioning smashing in the post for that puzzle. Might not have occurred to try that if it wasn't.
Got a solution for #4. The cascade of realizations about its hurdles was something else. I have a slightly improved version I could submit, but I want to try 1 other thing which might lead to greater gains.
E: Ended up worse.
I ended up working on #4 some more, instead of making a 15 input solution for #3. Even though I didn't like it at first, I like that it's the first puzzle here that doesn't have wildly branching solution paths. No smashing, state machines or special conditions. Just make the product and figure out what to do with the waste.
#5 really is a slog. Not done yet, but I'm on the right track. It just takes so long, even at speed 4, to get to the point where you can work out the next step.
As for #5, it doesn't seem too bad to me at first glance. You could just create a copy of the puzzle, substituting the Meitnerium for some other 6-bond atom like Iron, to be able to handle the construction part. Fusing upwards to Meitnerium though... yeah, that's where a lot of efficiency could be lost or won. Lots of flippity floppity math possible here.
E: Just realized what I initially had in mind is physically impossible. New track makes much more sense.