Loafer Synthesis
Loafer Synthesis
It's like glider synthesis but using loafers instead of gliders. Post your loafer syntheses here!
Re: Loafer Synthesis
GUYTU6J reported this post, saying that "it has been demonstrated that glider syntheses can be converted to loafer syntheses," and suggesting that the Sandbox would be a better place for this thread.
I don't entirely disagree -- but on the other hand, it doesn't seem like a terrible topic for the Patterns board.
In general, I'd like to ask everyone to go ahead and solve their own problems as far as where posts are supposed to go. If you feel strongly that there should be a thread in the Sandbox about loafer syntheses, well, go ahead and create a Sandbox thread for loafer syntheses.
Then have a discussion, if it turns out a discussion is needed. Depending on who agrees or disagrees, maybe one thread or the other will eventually get marked for deletion. Moderators can delete posts very easily, but the process of moving posts from one board to another takes some fiddling around... so when you can do it, please save the moderators' time for things like fighting spammers.
-
- Posts: 841
- Joined: June 27th, 2009, 10:58 am
- Location: Germany
Re: Loafer Synthesis
My 50 cent:
While the post cited above shows how a clean glider can be synthesized through loafers, and thus all pattern that can be glider constructed can also be loafer constructed, this doesn’t solve the question how to do this efficiently.
Thus I cannot see why this is not a valid topic for the Patterns board, provided the thread creator provides a nontrivial synthesis example or at least reasonable progress towards one. Then this is just another special interest thread like so many we already have, and rightfully. Diversity of interest is one of the fundamental drivers of progress.
I admit we need some discipline, and I strongly appreciate the efforts of the board admins and don’t want to put unnecessary workload on them, but some degree of freedom to think differently is also a prerequisite for a functioning community.
While the post cited above shows how a clean glider can be synthesized through loafers, and thus all pattern that can be glider constructed can also be loafer constructed, this doesn’t solve the question how to do this efficiently.
Thus I cannot see why this is not a valid topic for the Patterns board, provided the thread creator provides a nontrivial synthesis example or at least reasonable progress towards one. Then this is just another special interest thread like so many we already have, and rightfully. Diversity of interest is one of the fundamental drivers of progress.
I admit we need some discipline, and I strongly appreciate the efforts of the board admins and don’t want to put unnecessary workload on them, but some degree of freedom to think differently is also a prerequisite for a functioning community.
Re: Loafer Synthesis
Just register to add my .05
Along these same lines, the point here is that all the syntheses effort to date have been done using just gliders. That's logical, gliders are ubiquituous, its properties are well known, and there's a lot of mechanisms involving gliders: conduits, reflectors, construction arms, you name it.
But it is entirely possible that using any other spaceship, just by itself or in mixed reactions, another set of syntheses are possible. Not only with loafers ( of course not taking into account that reduction to glider by the Loafer -> glider reaction ) , but with *WSSs, SirRobins - doubtful - , copperheads... just by itself or in the company of others, including gliders. And the thread would be something like "UNORTHODOX SYNTHESES".
Nice forum, by the way. In Golly for many years.
Along these same lines, the point here is that all the syntheses effort to date have been done using just gliders. That's logical, gliders are ubiquituous, its properties are well known, and there's a lot of mechanisms involving gliders: conduits, reflectors, construction arms, you name it.
But it is entirely possible that using any other spaceship, just by itself or in mixed reactions, another set of syntheses are possible. Not only with loafers ( of course not taking into account that reduction to glider by the Loafer -> glider reaction ) , but with *WSSs, SirRobins - doubtful - , copperheads... just by itself or in the company of others, including gliders. And the thread would be something like "UNORTHODOX SYNTHESES".
Nice forum, by the way. In Golly for many years.
Re: Loafer Synthesis
Adding my £0.02:
I like the theoretical question of whether it's possible to construct things with loafers. But doing so efficiently is a less interesting problem to me because it's unlikely to be useful in any larger construction, and we still have a lot of other kinds of Optimization Game that actually might be useful. Of course I have no objection to people using this forum to do things that I find less interesting.
One similar project that would be useful is to find good single channel syntheses using LWSSes. These could be used to make a Speed Orthogonoid. (Or perhaps MWSSes would be better? That's what the original Orthogonoid used.)
I like the theoretical question of whether it's possible to construct things with loafers. But doing so efficiently is a less interesting problem to me because it's unlikely to be useful in any larger construction, and we still have a lot of other kinds of Optimization Game that actually might be useful. Of course I have no objection to people using this forum to do things that I find less interesting.
One similar project that would be useful is to find good single channel syntheses using LWSSes. These could be used to make a Speed Orthogonoid. (Or perhaps MWSSes would be better? That's what the original Orthogonoid used.)
Re: Loafer Synthesis
Veering off-topic a little further for my $.20 (interesting how for many people, inflation has apparently affected the old phrase "my two cents' worth")...
There are two or three possible universal-constructor projects that need a *WSS collision database. MWSSes would definitely be better for finding an orthogonal construction arm, presumably with recipes that can fire in eight possible directions, just like the current glider-based single-channel elbows. We don't actually have a reasonable slsparse-constructible Herschel-to-LWSS converter with a single-channel-compatible repeat time. One could be built easily enough, but it would be much bigger and more awkwardnessful than the H-to-MWSS that the Orthogonoid uses.
Besides that, we could also do the research and build a database for slow salvos made of *WSSes crashing into a faraway block, and see if construction costs end up being competitive with slow-salvo gliders. We know that slow *WSS salvos are universal, because at worst we could use a slow^2 elbow block, which we move around with *WSSes and fire gliders to do the actual construction. It would be relatively easy to put together a universal slow^2 elbow toolkit, but it hasn't been done yet.
... Okay, apologies for the tangent. Now, with any luck, back to our regularly scheduled topic of loafer syntheses --
Re: Loafer Synthesis
To add my 8 Dogecoin:
It's quite interesting that when someone says "a X synthesis", it can mean either:
In particular, of the two interactions below, which one is a glider synthesis and which one is a loafer synthesis?
Anyway, I agree entirely with Macbi's post. To add to that, single-channel *WSS technology is pretty much entirely what's needed to have a working demonstration of arbitrary rational velocities (a,b)c/p satisfying 2(a + b) < p. (Maybe the first (2,1)c/7 spaceship won't be found by ikpx2 after all...)
It's quite interesting that when someone says "a X synthesis", it can mean either:
- a synthesis of object X;
- a synthesis using copies of object X;
In particular, of the two interactions below, which one is a glider synthesis and which one is a loafer synthesis?
Code: Select all
x = 158, y = 59, rule = B3/S23
11bo2b2o3bo$11bo2bo3b2o$12bobo3bo$13bo5bo2$17b2o$16bo2bo$17bobo$18bo6$
151bo$151bobo$151b2o7$124bo$125b2o$124b2o2$128b3o$130bo$129bo2$131b3o$
131bo$132bo19b2o$151b2o$153bo3$156b2o$148b2o5b2o$147b2o8bo$149bo$124b
2o$123bobo$2o123bo$2bo$3bo$3o$o$6bo$5bobo16bo$b2o2bo2bo14b2o$2obo2b2o
14bo$23b3o$25bo$19bo$18bobo$17bo2b2ob3o$18b2o4b2o!
What do you do with ill crystallographers? Take them to the mono-clinic!
Re: Loafer Synthesis
Exactly. It's much more efficient to synthesize a pulsar with 2 loafers than 6.HartmutHolzwart wrote: ↑September 27th, 2020, 6:07 amMy 50 cent:
While the post cited above shows how a clean glider can be synthesized through loafers, and thus all pattern that can be glider constructed can also be loafer constructed, this doesn’t solve the question how to do this efficiently.
Thus I cannot see why this is not a valid topic for the Patterns board, provided the thread creator provides a nontrivial synthesis example or at least reasonable progress towards one. Then this is just another special interest thread like so many we already have, and rightfully. Diversity of interest is one of the fundamental drivers of progress.
I admit we need some discipline, and I strongly appreciate the efforts of the board admins and don’t want to put unnecessary workload on them, but some degree of freedom to think differently is also a prerequisite for a functioning community.
Code: Select all
x = 19, y = 9, rule = B3/S23
2obo2b2o3b2o2bob2o$b2o2bo2bobo2bo2b2o$5bobo3bobo$6bo5bo$o17bo$3o13b3o$
3bo11bo$2bo13bo$2o15b2o!
-
- Posts: 5142
- Joined: August 31st, 2020, 5:58 pm
Re: Loafer Synthesis
First of all, welcome to the forums.
Loafers are slow, so devoting time, effort, and resources to research loafer-based construction would be impractical. On the other hand, universal uni-directional XWSS construction is horribly inefficient, so if you're interested in alternative spaceship construction, you can look into that.
Speaking of loafer synthesis, why does the loafer require eight gliders? A loaf can be constructed with two gliders, and P-sparks are common, so the top part should be synthesizable with four gliders. Here's an encouraging sign:
Loafers are slow, so devoting time, effort, and resources to research loafer-based construction would be impractical. On the other hand, universal uni-directional XWSS construction is horribly inefficient, so if you're interested in alternative spaceship construction, you can look into that.
Speaking of loafer synthesis, why does the loafer require eight gliders? A loaf can be constructed with two gliders, and P-sparks are common, so the top part should be synthesizable with four gliders. Here's an encouraging sign:
Code: Select all
x = 10, y = 7, rule = B3/S23
2bo$obo$b2o2$2b3o2b3o$2bo4bo$3bo4bo!
I am tentatively considering myself back.
Re: Loafer Synthesis
I think it's a fairly common opinion that there's probably a six- or seven-glider loafer synthesis Out There Somewhere, but the usual avenues have been tried and nothing better than 8G has turned up -- yet.MathAndCode wrote: ↑September 27th, 2020, 1:12 pmSpeaking of loafer synthesis, why does the loafer require eight gliders?
It does seem like there ought to be a cheap way to build the loaf first and add the P spark later, but if I remember right, the last time we looked nothing too interesting showed up.
So a 4G-7G loafer synth wouldn't be building all the parts separately, it would be an all-at-once lucky creation of the entire loafer. That might well exist, but even an exhaustive 4G search is too big to be practical, and 5-7G is pretty far out of reach, until someone figures out a clever new line of attack.
-
- Posts: 5142
- Joined: August 31st, 2020, 5:58 pm
Re: Loafer Synthesis
Yes. Even a seven-glider synthesis is incompatible with building the three parts separately because the only two-glider synthesis of a loaf has a spark that prevents the glider collision for the P-spark from happening at the same time, none of the two-glider collisions for a P-spark have the right clearance for being done after the loaf has been formed, and we'd know if the third component could be synthesized with two gliders. The synthesis for the bottom component throws out a domino spark in the right direction to help five generations before the loafer is completed, but accessing it in the right way would be tough. An alternative three-glider collision for the bottom part that throws off different sparks might be helpful, but trying to create the bottom part with no degree of separation doesn't seem worth it until after at least two more decades of Moore's law.dvgrn wrote: ↑September 27th, 2020, 5:27 pmI think it's a fairly common opinion that there's probably a six- or seven-glider loafer synthesis Out There Somewhere, but the usual avenues have been tried and nothing better than 8G has turned up -- yet.MathAndCode wrote: ↑September 27th, 2020, 1:12 pmSpeaking of loafer synthesis, why does the loafer require eight gliders?
It does seem like there ought to be a cheap way to build the loaf first and add the P spark later, but if I remember right, the last time we looked nothing too interesting showed up.
So a 4G-7G loafer synth wouldn't be building all the parts separately, it would be an all-at-once lucky creation of the entire loafer. That might well exist, but even an exhaustive 4G search is too big to be practical, and 5-7G is pretty far out of reach, until someone figures out a clever new line of attack.
By the way, the current glider synthesis of the bottom part looks like it might have oscillator potential.
I suppose that I'm also not the first person to think of the following:
- Sir Robin is large and elongated, which suggests a partial-based synthesis.
- Sir Robin consists largely of space dust, which suggests that glider syntheses of it would likely have a greater proportion of gliders in the activation stage compared to other glider syntheses of spaceships of comparable size (or, more likely, extrapolations of the proportion of gliders in the activation recipe of other-sized spaceships to spaceships of that size).
- Points #1 and #2 suggest having a computer try to find the activation stage.
- Point #1 and Sir Robin's maximal speed for its direction suggest having the computer start with the back part.
Edit: Is there at least an eight-glider loafer synthesis where the loaf can be constructed arbitrarily long before the other parts?
I am tentatively considering myself back.