wwei23 wrote: October 16th, 2020, 12:09 am
But you can insert a domino while interacting with the pattern in a way that still allows the MWSS to be created.
I was very pessimistic, but JLS gives it a chance. No idea if synthesis is feasible, but there are numerous similar sparks so it seems there's a chance.
x = 15, y = 11
4bo5bo$5obobob5o$6b3o$$3boo5boo$bbobbo4boo$bbobbo$3booboo$5bobbo$5bobbo$6boo!
The 5S project (Smallest Spaceships Supporting Specific Speeds) is now maintained by AforAmpere. The latest collection is hosted on GitHub and contains well over 1,000,000 spaceships.
Semi-active here - recovering from a severe case of LWTDS.
Is there a pentadecathlon predecessor that evolves into another pentadecathlon predecessor?
Are there any Unit Non Totalistic Cellular Automaton Cells in Life?
Are Unit Wireworld Cells Possible?
Trump {Bot} wrote: October 18th, 2020, 7:57 pmIs there a pentadecathlon predecessor that evolves into another pentadecathlon predecessor?
Are there any Unit Non Totalistic Cellular Automaton Cells in Life?
Are Unit Wireworld Cells Possible?
1. Yes. Every n-generation pentadecathlon predecessor (n>0) evolves into an n-1-generation pentadecathlon predecessor.
2. Yes. Since Life is universal, all CAs that can be emulated by any computer can be emulated in Life. I don't know if anyone has actually constructed ones for non-totalistic CAs yet, but it's definitely possible.
3. Yes. See 2.
mniemiec wrote: October 18th, 2020, 8:15 pm2. Yes. Since Life is universal, all CAs that can be emulated by any computer can be emulated in Life. I don't know if anyone has actually constructed ones for non-totalistic CAs yet, but it's definitely possible.
The 0E0P metacell has that capability, although it is only designed for two-state rules (as opposed to Wireworld rules).
mniemiec wrote: October 18th, 2020, 8:15 pm2. Yes. Since Life is universal, all CAs that can be emulated by any computer can be emulated in Life. I don't know if anyone has actually constructed ones for non-totalistic CAs yet, but it's definitely possible.
The 0E0P metacell has that capability, although it is only designed for two-state rules (as opposed to Wireworld rules).
x = 1, y = 1, rule = MAPQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
o!
... you'll get a rule that shows exactly what an 0E0P metacell would do if programmed with that particular non-totalistic rule.
It takes a little more work (i.e., calls to lifelib) to actually build the Life pattern for that correctly-programmed 0E0P metacell, and then you can't run it through even one full metatick in a reasonable amount of time anyway. But technically this is a complete answer to the question, in practice and not just in theory.
Wasn't Gustavo's conjecture that the Caterpillar could be synthesized in less than 386 gliders, assuming it could be synthesized at all, the original inspiration for making a reverse caber-tosser?
wwei23 wrote: October 18th, 2020, 9:30 pm
Wasn't Gustavo's conjecture that the Caterpillar could be synthesized in less than 386 gliders, assuming it could be synthesized at all, the original inspiration for making a reverse caber-tosser?
It definitely could have been the inspiration. However, contrary to popular belief, Gustavo did not actually invent the reverse caber-tosser. He obviously did not invent—or even understand—the supposed universal constructor because:
He claimed that the information was encoded in the distance to a glider instead of some moving glider reflector. (It would be theoretically possible to make a universal constructor based on the former by making a binary counter then augmenting it until the glider arrived, but it would be much more work.) Absent of some unnecessarily elaborate mechanism, one glider arrival cannot carry enough information by itself for universal construction.
Gustavo claimed that the Gemini breeder required extra gliders because it was so big. This is contrary to the principle of how universal construction works. If nothing else, use a universal constructor with one glider to synthesize a universal constructor with fifteen gliders.
The other aspects of the thread, such as: the fact that it was apparently possible for Gustavo to receive Morse code signals from Massachusetts obviously not intended for him in southern Brazil; the claim of a secret leak being reminiscent of many false conspiracy theories; claiming that an elementary spaceship and a constructed spaceship could share a tagalong; various other outlandish details; implying that 385+15=390+1; and Gustavo's spammy nature in general, suggest that he had virtually no idea what he was talking about.
The idea of the information being stored in the distance to the final glider was originally proposed by calcyman (Adam Goucher), probably in collaboration with people at MIT, not Gustavo.
If the thread was even true, then the reverse caber-tosser would have been invented by Ivan Fomichev (codeholic), not Gustavo.
If anyone invented the RCT-based universal constructor in that thread, it was calcyman (as per point #4), although I would argue that that should not count as the actual invention (as per point #1).
Trump {Bot} wrote: October 18th, 2020, 7:57 pmIs there a pentadecathlon predecessor that evolves into another pentadecathlon predecessor?
Are there any Unit Non Totalistic Cellular Automaton Cells in Life?
Are Unit Wireworld Cells Possible?
1. Yes. Every n-generation pentadecathlon predecessor (n>0) evolves into an n-1-generation pentadecathlon predecessor.
2. Yes. Since Life is universal, all CAs that can be emulated by any computer can be emulated in Life. I don't know if anyone has actually constructed ones for non-totalistic CAs yet, but it's definitely possible.
3. Yes. See 2.
Thank you for answering.
For Donald Trump's question, here's an example:
ColorfulGalaxy wrote: October 19th, 2020, 5:09 pm
Could this be made into a pentadecathlon hassler?
With the appropriate guns, I don't see why not. The only problem is that there aren't true-period guns at P21, so you'll have to use a pseudo-period gun.
testitemqlstudop wrote:What do the three numbers after yl in an apgcode mean? (ylxxx_xxx_xxx_SEEMINGLYRANDOMCHARACTERS)
I think the first is "period"
I know another represents period of debris. Not sure which one.
But yes, the first is definitely period.
Perhaps one number encodes whether it’s SS linear growth (like blockstacker), SM (a gun), MS (a puffer), or MM (a rake)? I’m not sure.
If I remember correctly, this is what each number means:
First: the period of the population growth of the pattern. Note that this is not always the same as the pattern's "true" period (see below).
Second: The overall period of the debris.
Third: The amount the population grows by each period (again, this is based on the period of the population plot, not the period of the actual pattern).
Fourth: Some sort of hashing function, I think.
The explanation made me puzzled upon checking my search results in Glimmering Garden ikpx2_stdin census (yes, LaundryPizza03, that unknown user happens to be myself). Consisting of a p240 cloverleaf and a c/4d tubstretcher, this pattern is classified as yl60_1_30:
GUYTU6J wrote: October 24th, 2020, 5:20 amThe explanation made me puzzled upon checking my search results in Glimmering Garden ikpx2_stdin census (yes, LaundryPizza03, that unknown user happens to be myself). Consisting of a p240 cloverleaf and a c/4d tubstretcher, this pattern is classified as yl60_1_30: ... Whereas this is a yl240_1_120: ...
When classifying growing patterns, apgsearch categorizes things by their populations only, and not their cell configurations. In your first example, in additional to the wick-stretcher, there is a P240 (mod 60) oscillator, whose population cycles every 60 generations. In the second example, the oscillator briefly perturbs the end of the wick causing a slight change in its population, once every 240 generations. I'm not sure why the third number is half of the first in these cases. The text string at the end is a compressed histogram of the populations at each step of one cycle.
mniemiec wrote: October 24th, 2020, 7:08 am
When classifying growing patterns, apgsearch categorizes things by their populations only, and not their cell configurations. In your first example, in additional to the wick-stretcher, there is a P240 (mod 60) oscillator, whose population cycles every 60 generations. In the second example, the oscillator briefly perturbs the end of the wick causing a slight change in its population, once every 240 generations. I'm not sure why the third number is half of the first in these cases. The text string at the end is a compressed histogram of the populations at each step of one cycle.
The tubstretcher addes 2 cells to the vessel every 4 generations, and 2/4 = 0.5, which may account for the halved number.
Do you mean that in the case of yl60 apgsearch recognized the wickstretcher and the oscillator as a whole? I feel like blaming the object separation then, since the mechanism does not misidentify a block-laying switch engine with a pentadecathlon (say) in the ash behind.
GUYTU6J wrote: October 24th, 2020, 8:57 amDo you mean that in the case of yl60 apgsearch recognized the wickstretcher and the oscillator as a whole? I feel like blaming the object separation then, since the mechanism does not misidentify a block-laying switch engine with a pentadecathlon (say) in the ash behind.
I'm not sure. Object separation tends to get a bit wonky whenever two objects are close enough that their areas of infuence overlap, even if the objects don't do so. This is especially obvious in cases where there are two spaceships (especially of higher periods) are several spaces apart and it's obvious that they can never interact, yet they are still treated as a single "object".
In the case of the block-laying switch-engine and the pentadecathlon, it may depend on how far away the pentadecathlon is from everything else.
Has anyone conjectured that it is possible to make any finite pattern in ConwayLife by starting with some arrangement of four cells then doing nothing else besides removing cells in successive generations? That conjecture seems likely to be true to me.
MathAndCode wrote: October 24th, 2020, 3:19 pm
Has anyone conjectured that it is possible to make any finite pattern in ConwayLife by starting with some arrangement of four cells then doing nothing else besides removing cells in successive generations? That conjecture seems likely to be true to me.
Could you give an example? Do you start with a pattern that doesn't die in one generation and you (the player) just kill cells each generation?
MathAndCode wrote: October 24th, 2020, 3:19 pm
Has anyone conjectured that it is possible to make any finite pattern in ConwayLife by starting with some arrangement of four cells then doing nothing else besides removing cells in successive generations? That conjecture seems likely to be true to me.
Could you give an example? Do you start with a pattern that doesn't die in one generation and you (the player) just kill cells each generation?
Here is a way to make each of the three seven-cell acorns.
There is not a particular reason that I started with the prepond. The prepond, the four four-cell P spark parents, the three four-cell traffic light predecessors, and the five four-cell beehive predecessors are all interconvertible.
Last edited by MathAndCode on October 24th, 2020, 8:34 pm, edited 1 time in total.
There is not a particular reason that I started with the prepond. The prepond, the four four-cell P spark parents, the three four-cell traffic light predecessors, and the four four-cell beehive predecessors are all interconvertible.
That's a strange conjecture. I mean it seems to me that you can manipulate patterns in any way that could lead to any other pattern. Is there a maximum 1 cell killed/per generation rule? What if you had to kill cells until you reached your target pattern? Also, how long have you been working on this?
MathAndCode wrote: October 24th, 2020, 3:19 pmHas anyone conjectured that it is possible to make any finite pattern in ConwayLife by starting with some arrangement of four cells then doing nothing else besides removing cells in successive generations? That conjecture seems likely to be true to me.
Of course, the simplest (and least interesting) answer to this conjecture is "absolutely not", since GoE patterns are finite, and can't evolve from any arrangement of bits, 4 bits or not, so let's modify this to "every finite repeating pattern" instead.
T tetromino evolves into 3x3 hollow block, and taking one or two bits away from that gives the Pi and R evolutions, so it should be possible to get quite a variety of outcomes by manipulating those. To get to "any finite pattern" would need some fancy hand-waving like the RCT universal constructor, and even then, you only have constructions of everything that is glider-constructible. To get from there to "every finite still-life", or worse, "every finite repeating pattern" would require much more hand-waving.
Schiaparelliorbust wrote: October 24th, 2020, 4:05 pmIs there a maximum 1 cell killed/per generation rule?
No.
Schiaparelliorbust wrote: October 24th, 2020, 4:05 pmWhat if you had to kill cells until you reached your target pattern?
I don't think that it would work, at least not starting with only four cells.
Schiaparelliorbust wrote: October 24th, 2020, 4:05 pmAlso, how long have you been working on this?
I only thought of it earlier today.
mniemiec wrote: October 24th, 2020, 4:11 pmOf course, the simplest (and least interesting) answer to this conjecture is "absolutely not", since GoE patterns are finite, and can't evolve from any arrangement of bits, 4 bits or not, so let's modify this to "every finite repeating pattern" instead.
Yes, but one can make a denser region of cells then simply remove cells to get a Garden of Eden.
mniemiec wrote: October 24th, 2020, 4:11 pmT tetromino evolves into 3x3 hollow block, and taking one or two bits away from that gives the Pi and R evolutions, so it should be possible to get quite a variety of outcomes by manipulating those. To get to "any finite pattern" would need some fancy hand-waving like the RCT universal constructor, and even then, you only have constructions of everything that is glider-constructible.
Why not just figure out a way to create arbitrarily large regions of all on cells?
