How does one start creating stuff here?

For general discussion about Conway's Game of Life.
Melanthius
Posts: 1
Joined: March 11th, 2015, 3:45 pm

How does one start creating stuff here?

Post by Melanthius » March 11th, 2015, 3:46 pm

Hi. I'm new to this forums and new to Conway's Game of Life, I just discovered it today and I found it very interesting, I, however, do not understand how it works. I've seen people creating cool stuff and I want to start creating cool stuff as well, thanks in advance for help.

User avatar
Alexey_Nigin
Posts: 323
Joined: August 4th, 2014, 12:33 pm
Location: Ann Arbor, MI
Contact:

Re: How does one start creating stuff here?

Post by Alexey_Nigin » March 11th, 2015, 4:02 pm

I have sent you a PM regarding this.
There are 10 types of people in the world: those who understand binary and those who don't.

User avatar
Kiran
Posts: 285
Joined: March 4th, 2015, 6:48 pm

Re: How does one start creating stuff here?

Post by Kiran » August 23rd, 2015, 11:23 am

Posts: 1
I see the result of your PM.
Kiran Linsuain

User avatar
Alexey_Nigin
Posts: 323
Joined: August 4th, 2014, 12:33 pm
Location: Ann Arbor, MI
Contact:

Re: How does one start creating stuff here?

Post by Alexey_Nigin » August 23rd, 2015, 2:23 pm

Kiran wrote:
Posts: 1
I see the result of your PM.
Humans are strange things; I must say I don't know how they work. At all.

I cannot understand a human who says "I want to learn", receives a reply "OK, I will teach you", and disappears forever.

However, I have to admit that such humans exist in great amounts.

This is like advertising apgsearch/wsearch. I spent quite a lot of time on it, and here is a usual conversation:

"Hello, do you know what CA are?"
"Yep. You read a small lecture about them long ago."
"Would you like to run (apg/w)search, which is ... You will find a lot of ..."
"Oh yes! This is so interesting! What should I do?"
"Give me your email."
"Here it is."

In email: "Hello, thanks for ... The program can be downloaded from ..."

Neverending silence. (Optionally preceded by a reply "I'll download it tomorrow")
There are 10 types of people in the world: those who understand binary and those who don't.

User avatar
Kazyan
Posts: 959
Joined: February 6th, 2014, 11:02 pm

Re: How does one start creating stuff here?

Post by Kazyan » August 23rd, 2015, 3:20 pm

It might be because people often don't want things handed to them. Conway's Game of Life holds a deep well of secrets to discover, and those who get tentatively interested in it seem to have a "Wow! I wonder what I can find?" attitude. This is great, but immediately throwing them into the world built by our 45 years of research probably curbs the initial enthusiasm.

I think it's good for a newbie Lifenthusiast to play around with doodles and methuselahs, and to stare at the files in Golly while wondering how in the world people found/made them. The "useless discoveries" thread seems to be a great acclimatization zone, perhaps because it's an unstructured but high-intensity learning area. gmc_nxtman, The Turtle, A for Awesome, and even myself all hung around there while learning the ropes.

Then, once we catch up and start to understand what everyone's doing these days, the research begins.

And I doubt that a structured tutorial would be more effective. Conway's Game of Life is ordered chaos--doesn't it make sense that the kind of people who get drawn to it would do better in an learning environment that is also ordered chaos? Ever notice how newcomers seem to skip the "Game of Life Links" sticky and try to find things out on their own?
Tanner Jacobi
Coldlander, a novel, available in paperback and as an ebook.

User avatar
The Turtle
Posts: 102
Joined: May 6th, 2015, 8:14 pm
Location: Chicago, Illinois

Re: How does one start creating stuff here?

Post by The Turtle » August 24th, 2015, 6:36 pm

Kazyan wrote:It might be because people often don't want things handed to them. Conway's Game of Life holds a deep well of secrets to discover, and those who get tentatively interested in it seem to have a "Wow! I wonder what I can find?" attitude. This is great, but immediately throwing them into the world built by our 45 years of research probably curbs the initial enthusiasm.

I think it's good for a newbie Lifenthusiast to play around with doodles and methuselahs, and to stare at the files in Golly while wondering how in the world people found/made them. The "useless discoveries" thread seems to be a great acclimatization zone, perhaps because it's an unstructured but high-intensity learning area. gmc_nxtman, The Turtle, A for Awesome, and even myself all hung around there while learning the ropes.

Then, once we catch up and start to understand what everyone's doing these days, the research begins.

And I doubt that a structured tutorial would be more effective. Conway's Game of Life is ordered chaos--doesn't it make sense that the kind of people who get drawn to it would do better in an learning environment that is also ordered chaos? Ever notice how newcomers seem to skip the "Game of Life Links" sticky and try to find things out on their own?
Well, I'm sure many people are turned down because they find things that have been discovered a long time ago. I would have left this site if, on my first post, I posted a R-pentomino relative, and someone replied: "Oh, that's an R-pentomio relative." However, what I posted on my first post was what simsim314 calls a "flipper", which no one said was discovered before. That may be the reason why I didn't quit on Conway's Game of Life. It is hard to catch up on what everyone is doing today.

Oh, and here's the flipper.

Code: Select all

x = 18, y = 21, rule = B3/S23
o$b2o$2o4$2bo$obo3bobo$b2o4b2o$7bo3$8bo$9bo6b2o$7b3ob2obo2bo$12bob3o$
12bo$11b2ob3o$12bobo2bo$10bobobobo$10b2o3bo!
Only two things are constant: change and the speed of light.

User avatar
biggiemac
Posts: 504
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: How does one start creating stuff here?

Post by biggiemac » August 25th, 2015, 2:03 am

It's very rare to join and immediately have something new to contribute, but it does happen. I played around with golly for a couple years on and off before joining, and only joined the forums to add my thoughts to the construction of the Waterbear spaceship. (As it turned out I stuck around to assemble the whole thing). Andrew Wade took everyone by surprise when he joined with the Gemini spaceship complete, orders of magnitude better than existing forum members' estimates.

People who join with questions like the above though seem to have a different process for figuring things out that doesn't lend well to joining right away with something new. I like to be set free to reverse-engineer old discoveries myself, but others prefer to learn with a walkthrough. Perhaps what would be best for them is to have a veteran forum member point them to a subfolder in Golly with lots to choose from and ask which of the options they find the coolest, and then reverse-engineer their choice together. At least with a single pattern to delve into it may be more exciting and less overwhelming.

No matter what, it's great to join with something to contribute besides enthusiasm. If one doesn't have ideas or something they feel is new, they can always contribute CPU hours! Apple Bottom comes to mind there, who wasn't a forum user until posing about an apgsearch patch that saved logs locally.
Physics: sophistication from simplicity.

User avatar
Apple Bottom
Posts: 1033
Joined: July 27th, 2015, 2:06 pm
Contact:

Re: How does one start creating stuff here?

Post by Apple Bottom » September 29th, 2015, 8:33 am

Since I'm a newcomer myself, as biggiemac observes, I thought I'd offer my perspective.

Conway's Game of Life is fascinating, but it's also utterly intimidating when you look at the state of the art as a beginner. The basics are easy enough to understand; and there's no real barrier to downloading a copy of Golly, drawing patterns, and seeing how they evolve. But the learning curve is incredibly steep after that, and going from the very basics to understanding the frontiers of current research - much less contributing on that level! - will appear almost impossible when you're just getting started.

Myself, I felt and still feel like a child who's barely learned how to write an "A", a "B" and so on, and who's now been thrust into a postdoc-level class on the subtler points of modern literature. Half of the time I don't even understand what other folks are saying. And when I do, I still couldn't hope to contribute anything worthwhile to the discussion myself.

I reckon I could grit my teeth and work my way up to that level. But honestly? I don't believe I'm bright enough for that.

Still, downloading and running apgsearch allows me to help out without knowing much. I can contribute the odd CPU-hour; and if there is something I can do, like scratch an itch of mine (local logging) and share the result, or if I just so happen to get lucky and stumble across something interesting (like that new yl1152_1_275_a3591abbc6ddbf461332c01353cae282 pattern), all the better.

Also - I'm still a reasonably technical user. The thought of opening a shell doesn't scare me. I know what a Makefile is. I'd reckon there's a lot of mathematically-inclined folks out there for who this ain't true, and having a binary package of an apgsearch version with a GUI that can easily be installed and run (especially on other operating systems, such as Mac OS X or Windows) would go a long way to reaching those folks.
If you speak, your speech must be better than your silence would have been. — Arabian proverb

Catagolue: Apple Bottom • Life Wiki: Apple Bottom • Twitter: @_AppleBottom_

Proud member of the Pattern Raiders!

User avatar
A for awesome
Posts: 2047
Joined: September 13th, 2014, 5:36 pm
Location: 0x-1
Contact:

Re: How does one start creating stuff here?

Post by A for awesome » September 29th, 2015, 8:46 am

Apple Bottom wrote:Also - I'm still a reasonably technical user. The thought of opening a shell doesn't scare me. I know what a Makefile is. I'd reckon there's a lot of mathematically-inclined folks out there for who this ain't true, and having a binary package of an apgsearch version with a GUI that can easily be installed and run (especially on other operating systems, such as Mac OS X or Windows) would go a long way to reaching those folks.
As one of those "mathematically-inclined folks out there for who this ain't true", I completely agree!
Apple Bottom wrote:Since I'm a newcomer myself, as biggiemac observes, I thought I'd offer my perspective.

Conway's Game of Life is fascinating, but it's also utterly intimidating when you look at the state of the art as a beginner. The basics are easy enough to understand; and there's no real barrier to downloading a copy of Golly, drawing patterns, and seeing how they evolve. But the learning curve is incredibly steep after that, and going from the very basics to understanding the frontiers of current research - much less contributing on that level! - will appear almost impossible when you're just getting started.

Myself, I felt and still feel like a child who's barely learned how to write an "A", a "B" and so on, and who's now been thrust into a postdoc-level class on the subtler points of modern literature. Half of the time I don't even understand what other folks are saying. And when I do, I still couldn't hope to contribute anything worthwhile to the discussion myself.
That is a real problem with the state of modern-day Life. It might be useful to have a "User's manual to Life" that contains all the information one needs to start contributing at a high level.
x₁=ηx
V ⃰_η=c²√(Λη)
K=(Λu²)/2
Pₐ=1−1/(∫^∞_t₀(p(t)ˡ⁽ᵗ⁾)dt)

$$x_1=\eta x$$
$$V^*_\eta=c^2\sqrt{\Lambda\eta}$$
$$K=\frac{\Lambda u^2}2$$
$$P_a=1-\frac1{\int^\infty_{t_0}p(t)^{l(t)}dt}$$

http://conwaylife.com/wiki/A_for_all

Aidan F. Pierce

User avatar
Apple Bottom
Posts: 1033
Joined: July 27th, 2015, 2:06 pm
Contact:

Re: How does one start creating stuff here?

Post by Apple Bottom » September 29th, 2015, 1:07 pm

Apple Bottom wrote:Also - I'm still a reasonably technical user. The thought of opening a shell doesn't scare me. I know what a Makefile is. I'd reckon there's a lot of mathematically-inclined folks out there for who this ain't true, and having a binary package of an apgsearch version with a GUI that can easily be installed and run (especially on other operating systems, such as Mac OS X or Windows) would go a long way to reaching those folks.
To elaborate further on this, GUIs could be kept completely separate from the current apgsearch by having the two communicate through standard input and output (and error) only.

This would keep apgsearch platform- and toolkit-independent, and allow Andrew to keep his attention focussed on the core functionality of soup-searching, without getting side-tracked. It'd also allow for different GUIs competing on equal footing. And it'd allow for a single GUI to control an arbitrary number of apgsearch instances.

(A note on instances: I know apgsearch can use OpenMP, but I prefer to fire up instances manually myself; it gives me more flexibility to free CPU cores when I need them for other things, while keeping them all busy when I'm away.)

The only change I'd possibly make to apgsearch itself is adding a mode where output to stdout is intended for machine rather than human consumption. But even that ain't a given.
A for awesome wrote:That is a real problem with the state of modern-day Life. It might be useful to have a "User's manual to Life" that contains all the information one needs to start contributing at a high level.
The Life wiki's a great start, perhaps that could be leveraged?
If you speak, your speech must be better than your silence would have been. — Arabian proverb

Catagolue: Apple Bottom • Life Wiki: Apple Bottom • Twitter: @_AppleBottom_

Proud member of the Pattern Raiders!

User avatar
dvgrn
Moderator
Posts: 6737
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: How does one start creating stuff here?

Post by dvgrn » September 30th, 2015, 8:46 am

Apple Bottom wrote:Myself, I felt and still feel like a child who's barely learned how to write an "A", a "B" and so on, and who's now been thrust into a postdoc-level class on the subtler points of modern literature. Half of the time I don't even understand what other folks are saying. And when I do, I still couldn't hope to contribute anything worthwhile to the discussion myself.
It's funny how familiar that sounds. I'm a newcomer to Conway's Life research myself... picked it up only in 2001, joining a group of maybe two or three dozen people, many of whom had been working on Life problems off and on for just about thirty years at that point.

After my first beginner's-luck discovery (the now-obsolete boojum reflector) I couldn't find much of anything else worth contributing for more than a year, and even after several years it seemed as if I was still on the newbie side of the bell curve. For that matter, there are still subtopics in this business that look like black magic to me.

However, it's perfectly possible to pick up useful expertise in one area that you're really interested in, or just a few areas. I like biggiemac's suggestion of picking a pattern for some kind of reverse-engineering walkthrough, to flatten out the learning curve.

That's easier said than done of course. For many complex patterns the important thing is not the particular set of magical-looking mechanisms that happen to work, but the much larger list of mechanisms that might work.

Building up that mental toolkit is the hard part. It won't necessarily help if someone just hands you a big stamp collection -- it's really good to have seen and recognized the pieces in action in other contexts.

Something I have on a back burner, now that Chris Rowett's amazing scriptable LifeViewer is available, is to try explaining the functioning of a complex pattern, using an animated flyover with pop-up labels. Anyone have a suggestion for a pattern to start with? I guess a metapixel would be a little too big, at least for a forum posting, unless Nathaniel wants to consider bumping up the message limit a little -- even the OFF state is 63K.
Apple Bottom wrote:
A for awesome wrote:That is a real problem with the state of modern-day Life. It might be useful to have a "User's manual to Life" that contains all the information one needs to start contributing at a high level.
The Life wiki's a great start, perhaps that could be leveraged?
Good idea. The problem is getting anyone to write down all that information in any kind of organized way. Solving Conway's Life puzzles is fun, but explaining the solutions clearly is hard work...!

For big projects that have been completed via collaboration on the forums, like the waterbear or centipede spaceships, a decent substitute for a formal walkthrough would be just reading from the top of the thread and patiently following through all the stages of development.

In the search-utility department, at least, there have been a number of attempts to bridge the information gap. I just put on my LifeWiki to-do list the idea of linking from search utilities' LifeWiki pages to any walkthroughs that have been written.

Can anyone think of other online explanations where people might benefit from a (clearly labeled) LifeWiki link?

User avatar
triller
Posts: 113
Joined: September 3rd, 2009, 2:41 pm

Re: How does one start creating stuff here?

Post by triller » September 30th, 2015, 11:59 am

dvgrn wrote:Can anyone think of other online explanations where people might benefit from a (clearly labeled) LifeWiki link?
That's a difficult nut to crack. I'll cite some examples:
David Buckingham's marvelous Adder (c. 1975) was a stock pattern that I acquired with my downloaded version of Winlife. The description block states size optimization by Mark Niemiec, but the original version (pre-optimization) isn't to be had.
My own constructions are liberally blathered on, but only in broad strokes. I've never used a search utility, but rather fit things up by brute force trial and error, and present only the completed package.
Even Andrew Wade's Gemini wasn't really developmentally described. That was more like 'here's a construction arm based spaceship, and oh, by the way it's a KNIGHTSHIP!'
biggiemac's regular posting of progress on the waterbear is a fine example of how an end goal was incrementally arrived at.
I guess each practitioner works a little differently, as is naturally human. The thing is that divine inspiration is a tough commodity to quantify. For myself it has most often been mere dumb luck.

Triller
The most exciting phrase to hear in science, the one that heralds new
discoveries, is not "Eureka!" (I found it!) but "That's funny ..."
-- Isaac Asimov

Moth-Wingthane
Posts: 12
Joined: October 11th, 2015, 8:54 pm

Re: How does one start creating stuff here?

Post by Moth-Wingthane » October 12th, 2015, 5:55 am

This is all very new to me and I'm only just getting my head around the very basic stuff still, hoping to contribute more than just CPU cycles in the future. But I have to say a big thanks to Apple Bottom for getting me up and running quickly! I'm relatively new to Linux as well, and it would have taken me quite some time to sort it out on my own, I'm sure :) (I didn't know there was a forum I could come and ask on either!)

User avatar
Alexey_Nigin
Posts: 323
Joined: August 4th, 2014, 12:33 pm
Location: Ann Arbor, MI
Contact:

Re: How does one start creating stuff here?

Post by Alexey_Nigin » October 12th, 2015, 11:24 am

Moth-Wingthane wrote:This is all very new to me and I'm only just getting my head around the very basic stuff still, hoping to contribute more than just CPU cycles in the future. But I have to say a big thanks to Apple Bottom for getting me up and running quickly! I'm relatively new to Linux as well, and it would have taken me quite some time to sort it out on my own, I'm sure :) (I didn't know there was a forum I could come and ask on either!)
Welcome on board!
There are 10 types of people in the world: those who understand binary and those who don't.

Dets65
Posts: 23
Joined: May 7th, 2013, 8:20 am

Re: How does one start creating stuff here?

Post by Dets65 » October 12th, 2015, 4:26 pm

I've been lurking here for a while and have NO idea how people make ANY of these things. I mean, I know that search programs exist, but how in the world was Gemini made? Or that massive p400-something gun that's included with Golly? Where people make stuff by hand, do they just brute force and make random combinations until they come up with something? I don't get it :( Can someone please help me actually MAKE something for once that's not some stupid, trivial glider synthesis of a block that's way more cumbersome than the normal one?

How were the Eaters and patterns like it made? People throwing together mildly symmetrical designs until something works on the 150th try? Or is that an extremely valid way of doing things and I'm just completely oblivious? Granted, I do understand some simple reactions. Should I try making small pieces with them and arranging that into something bigger? Or is CPU time better than human time (constructing and experimenting with reactions and stabilizing them into something bigger) by this point, and I should just download APGsearch and leave it running?

User avatar
dvgrn
Moderator
Posts: 6737
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: How does one start creating stuff here?

Post by dvgrn » October 12th, 2015, 5:18 pm

Dets65 wrote:I've been lurking here for a while and have NO idea how people make ANY of these things. I mean, I know that search programs exist, but how in the world was Gemini made? Or that massive p400-something gun that's included with Golly?
I'm not sure which p400-something gun you're referring to, but the Gemini would make a great reverse-engineering case study. The idea of a walkthrough, or Q&A session, or something, was previously mentioned on this thread, but I'm not sure how it would work exactly.

I could quickly generate a lot of text about the Gemini, because of getting to know it in fairly fine detail from having to make adjustments to build the Gemini guns. But if history is any guide, that might end up being too long and boring for anyone to read...!

One possible place to start is some short explanatory writeups from back when the Gemini suddenly appeared on the scene. Does any of that bring up more questions?
Dets65 wrote:Where people make stuff by hand, do they just brute force and make random combinations until they come up with something? I don't get it :( Can someone please help me actually MAKE something for once that's not some stupid, trivial glider synthesis of a block that's way more cumbersome than the normal one?
That's a tricky request. Gustavo has been wrestling with the same problem recently. "Making stuff" isn't just a matter of brute force, but there's certainly a lot of luck involved.

You have to more or less do what Conway Himself talked about, way back -- just spend time trying out all the various mechanisms until you can recognize them in the context of larger patterns, and "live with" the rule for a while to get a sense of the possibilities.

You usually can't just throw CPU cycles at problems, because search spaces can very easily be made so big that you never get through them... and unless your search is really cleverly designed, if you only get through 0.01% of a search, probably you're repeating a lot of trivial permutations in some boring corner of the search space, and there won't be anything interesting there.

You have to do kind of a delicate balancing act, setting up searches that are challenging but not quite impossible for your particular hardware. That way you're most likely looking in a search space that nobody has ever looked in before -- and if there happens to be something out there, then you'll find it.

User avatar
biggiemac
Posts: 504
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: How does one start creating stuff here?

Post by biggiemac » October 12th, 2015, 5:22 pm

Edit: I started writing before dvgrn's post went up so I haven't scanned for commonalities. If any of this is redundant, oh well.

Large engineering usually involves turning the overwhelming number of possibilities into a smaller, controlled flow of interestingness. Typically one has to start with lucky/brute-forced/magic findings to make the building blocks: (these gliders plus this block equals this, etc), and then once enough building blocks are available, putting things together.

The self-constructing circuitry, for example, took known spartan Herschel conduits and known interactions between a glider salvo and a block and engineered a way to translate between a tape and an outgoing glider. That's the flow of interestingness: the information on the tape goes into the Herschel conduits with specific timing to generate the salvo that hit the block to position and fire the glider. That made a bigger building block, where now someone could take the tape->glider component and a set of known slow salvo recipes and build something. In Gemini, one further stroke of insight removed the tape and fed the Hershel conduits directly, allowing a self-constructing pattern smaller than anyone expected. But this is only something anyone could have done thanks to prior work, thanks to earlier efforts to chart flows of interestingness more controllable than "born with 3 neighbors, surviving with 2 or 3."

Someone can contribute at any level, but generally the intuition-driven approach to engineering only really pays off if you are taking known pieces and coordinating them into something bigger. The super-interesting little parts (syringe, loafer, snark) all come more from a computational approach, where the human work is spent making effective software.

There are also ways to start from a big goal and set up a search for little parts that can aid it, but that requires some statistical certainty that such a part exists to ever be fruitful. That was a part of the waterbear construction - nobody had had general reason to keep track of Herschel flyby reactions that turned gliders into beehives until I saw that it could make an adjustable track in this very-specific caterpillar construction, and given the commonness of such a reaction it was easy to find that it existed.

Hope this comment (which turned into a pretty open ended ramble) gives people some new and helpful information.
Physics: sophistication from simplicity.

Dets65
Posts: 23
Joined: May 7th, 2013, 8:20 am

Re: How does one start creating stuff here?

Post by Dets65 » October 12th, 2015, 5:50 pm

I don't want to reply to both comments, but thank you for the info! That cleared that up. When you look at the Gemini, it seems so daunting, but then you realize... "Oh wait, that's a glider... And that's a Herschel! Oh, looky here! There's a block and an eater too!" I'm not currently on the level of reading about exactly how every piece of the Gemini works. Guns are pretty simple (in concept) when you really think about it; Just a bunch of gliders being precisely ordered and reflected so that they synthesize a ship from a group of still-lifes, for example. All you have to do is arrange the reflectors and such precisely to match up with the known syntheses. Finding syntheses might be a job best left to the computers, haha. Granted, sometimes it is pretty obvious which still-lifes are a part of a ship or can quickly evolve into one once the gliders get into the mix.

Sorry for the comparatively very short reply, and a big, hearty thank you to both of you for setting me on the right track...

Just don't expect an optimized Weekender gun coming from me... By, well, the weekend. :lol:

(When I said I was talking about the p400 gun, I meant this one in the built in pattern collection of Golly. http://conwaylife.com/wiki/P416_60P5H2V0_gun )

Moth-Wingthane
Posts: 12
Joined: October 11th, 2015, 8:54 pm

Re: How does one start creating stuff here?

Post by Moth-Wingthane » October 12th, 2015, 8:20 pm

Alexey_Nigin wrote: Welcome on board!
Thanks! ^^

User avatar
biggiemac
Posts: 504
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: How does one start creating stuff here?

Post by biggiemac » October 12th, 2015, 9:37 pm

Dets65 wrote:Finding syntheses might be a job best left to the computers
While there are some programs that perform a random glider collision, they haven't been the source of most syntheses. The synthesis techniques were what I had in mind when I wrote "magic" in my earlier reply, because they still tend to feel like magic to me.

It can be roughly broken into two parts: 1) make something that works, and 2) minimize the number of gliders. I sometimes feel capable at the latter, but the former requires quite the arsenal of "X number of gliders can make this component or spark or what have you," as well as making a path for each glider that doesn't cause premature interaction. Some people have quite the knack for it (mniemiec and Extrementhusiast come to mind), but even they are largely drawing upon building blocks found by computers or by people before them and supplying the necessary insight.

I'm glad our comments helped you understand something new! :D Feel free to keep asking things, I like how this thread has turned into a place where beginners and less-beginners can share thoughts on just "how does one start creating stuff here."
Physics: sophistication from simplicity.

Dets65
Posts: 23
Joined: May 7th, 2013, 8:20 am

Re: How does one start creating stuff here?

Post by Dets65 » October 12th, 2015, 10:23 pm

biggiemac wrote:
Dets65 wrote:Finding syntheses might be a job best left to the computers
While there are some programs that perform a random glider collision, they haven't been the source of most syntheses. The synthesis techniques were what I had in mind when I wrote "magic" in my earlier reply, because they still tend to feel like magic to me.

It can be roughly broken into two parts: 1) make something that works, and 2) minimize the number of gliders. I sometimes feel capable at the latter, but the former requires quite the arsenal of "X number of gliders can make this component or spark or what have you," as well as making a path for each glider that doesn't cause premature interaction. Some people have quite the knack for it (mniemiec and Extrementhusiast come to mind), but even they are largely drawing upon building blocks found by computers or by people before them and supplying the necessary insight.

I'm glad our comments helped you understand something new! :D Feel free to keep asking things, I like how this thread has turned into a place where beginners and less-beginners can share thoughts on just "how does one start creating stuff here."
It does kind of feel like magic, haha. But sometimes you look at a still life and go "Wow, that looks just like this spaceship part." Usually you only realize that when looking at it being built in a gun, though. Thanks for the clarification! I feel like I can actually look at patterns now without my mind imploding from wondering how they did it. Also, it's great to hear from someone who you know is fairly trusted. I mean, you made the waterbear, after all! (Just one thing I have to ask; Did you place all of the cells manually, and if you did, do you have carpal tunnel? :lol: ) I don't think you overlapped much with dvgrn's post. If I have any questions, at least I know where to ask :) Have to go to bed. Will tinker with some things tomorrow.

Once again, a big, hearty THANK YOU (for both of you) for all of the stuff you have done to get me on the right track. I hope that if someone else is coming across this in the future, it helps them as well. Honestly, you didn't need to devote all of this time to me. But you did, and that is why I commend you both. :D

EDIT; Just one question. How are huge creations like the waterbear and gemini made? Is it mostly copy-pasted, or is it all hand placed? It baffles me both the scale of these things and how sometimes only one person is making it, and also the fact that if, say, one glider is erased the whole thing breaks.

User avatar
biggiemac
Posts: 504
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: How does one start creating stuff here?

Post by biggiemac » October 12th, 2015, 11:01 pm

The Waterbear was mostly copy-pasted; in 79 generations the H climber would go through its full cycle so all I had to do was figure out where the next H had to go and copy and paste a glider track and an H in the right phase from somewhere else on the ship. And again it relied on building blocks. Once there was a tracks-to-eastward-LWSS contraption worked out, I copy-pasted that whole thing and matched it to the spot it belonged whenever possible.

I think a lot of other large designs are assembled by a script (HBKs, Centipede, Caterpillar, I think also Gemini) that is fed the building blocks and what to do to assemble them. With the Waterbear I just sort of went for a learning-while-doing approach for what components would be wise to make, and by the time I had made enough to maybe feed into a script I was close enough to done that I finished it by hand.

So for large patterns it goes something like
With the pencil tool: goodness, no.
Copy-pasting: yes at the beginning, sometimes the whole way through.
Computer-aided: if it seems doable, this is the way to get something completed. (Though it may not be optimal)
Physics: sophistication from simplicity.

Dets65
Posts: 23
Joined: May 7th, 2013, 8:20 am

Re: How does one start creating stuff here?

Post by Dets65 » October 13th, 2015, 12:26 pm

biggiemac wrote:The Waterbear was mostly copy-pasted; in 79 generations the H climber would go through its full cycle so all I had to do was figure out where the next H had to go and copy and paste a glider track and an H in the right phase from somewhere else on the ship. And again it relied on building blocks. Once there was a tracks-to-eastward-LWSS contraption worked out, I copy-pasted that whole thing and matched it to the spot it belonged whenever possible.

I think a lot of other large designs are assembled by a script (HBKs, Centipede, Caterpillar, I think also Gemini) that is fed the building blocks and what to do to assemble them. With the Waterbear I just sort of went for a learning-while-doing approach for what components would be wise to make, and by the time I had made enough to maybe feed into a script I was close enough to done that I finished it by hand.

So for large patterns it goes something like
With the pencil tool: goodness, no.
Copy-pasting: yes at the beginning, sometimes the whole way through.
Computer-aided: if it seems doable, this is the way to get something completed. (Though it may not be optimal)
Interesting! Also, what exactly is an "H Climber"? I tried searching it and can't find anything on google :|

After looking up more information on the Waterbear and seeing it in action, WOW. That is an amazing design. Especially in the fact that there are only two base principles that run it.
1. The hershels producing two gliders on different, but similar, paths,
and 2. Glider synthesis, both of the fuel helix and the still lifes that end up eating the gliders.
I just find it amazing that the Hershels are multipurpose, both being the source of the obliqueness and also being a means to duplicate gliders to provide more for synthesis. Just a purely ingenious design.

User avatar
biggiemac
Posts: 504
Joined: September 17th, 2014, 12:21 am
Location: California, USA

Re: How does one start creating stuff here?

Post by biggiemac » October 13th, 2015, 1:51 pm

H climber = Herschel climber, the base oblique reaction of the Waterbear; sorry for any confusion :)

By the time I joined the forums, the oblique climbing reaction that generated two gliders was long known, and the helix with the correct velocity had been engineered by codeholic's script. Those building blocks gave me a starting point. The design bits I added included coming up with a group of tracks that could provide universal construction, using that to do the syntheses, and engineering the "resets" that made the ship three medium triangles instead of one enormous triangle (I think dvgrn's estimate was that the ship would be about 20 times its current size if there weren't resets).

Caterpillar designs like the Waterbear are made by taking any small pattern that is multipurpose (like the Herschel in this context), and using their versatility to support a self-sustaining loop. The Pis in the original Caterpillar are also multipurpose: they set the velocity, generate the gliders, and can even rephase passing gliders and blinkers. So any time someone finds a small reaction that moves and gives back everything it started with and more, that's a possible starting point for a caterpillar. There are a couple of other reactions that work ((13,1)c/31 with a different Herschel reaction is one, an HBK-caterpillar using the reaction in codeholic's avatar is another) but they are expected to be a lot more work to complete.
Physics: sophistication from simplicity.

Dets65
Posts: 23
Joined: May 7th, 2013, 8:20 am

Re: How does one start creating stuff here?

Post by Dets65 » October 13th, 2015, 3:42 pm

biggiemac wrote:H climber = Herschel climber, the base oblique reaction of the Waterbear; sorry for any confusion :)

By the time I joined the forums, the oblique climbing reaction that generated two gliders was long known, and the helix with the correct velocity had been engineered by codeholic's script. Those building blocks gave me a starting point. The design bits I added included coming up with a group of tracks that could provide universal construction, using that to do the syntheses, and engineering the "resets" that made the ship three medium triangles instead of one enormous triangle (I think dvgrn's estimate was that the ship would be about 20 times its current size if there weren't resets).

Caterpillar designs like the Waterbear are made by taking any small pattern that is multipurpose (like the Herschel in this context), and using their versatility to support a self-sustaining loop. The Pis in the original Caterpillar are also multipurpose: they set the velocity, generate the gliders, and can even rephase passing gliders and blinkers. So any time someone finds a small reaction that moves and gives back everything it started with and more, that's a possible starting point for a caterpillar. There are a couple of other reactions that work ((13,1)c/31 with a different Herschel reaction is one, an HBK-caterpillar using the reaction in codeholic's avatar is another) but they are expected to be a lot more work to complete.
You know what they say- "If I have seen further, it is by standing on the shoulders of giants." :) It is a very impressive design indeed (I could certainly not make it, haha. The timing is too perfect, the theory is too complex (at times, though I can understand much of it), and the scale is too massive), but to only congratulate you would be to undermine several others, including Codeholic, as well as those who researched the Herschel and the glider syntheses to find the reactions needed. Most life patterns made by one person are just the realization that other's work can be combined into a larger pattern. (Aside from computer-found patterns, of course)

To illustrate the pattern's complexity, I tried removing one block in the lower triangle. (Specifically, a block from a glider) This lack of a single glider destroyed more gliders, ruining the synthesis, and making the fuel helix useless. Of course, it took a while for the broken part of the helix to reach the top (The helix was not fully broken, a part was just shaved off) but the main glider-producing mechanism went haywire, ultimately destroying the entire ship.

Post Reply