Dependent Reflector Terminology and Analysis

[Jormungant]

... Let's see if I am stepping on a land mine here.

A signal is not necessarily a glider or a xwss indeed, and can take many form, but in essence a "signal" should have some entropy, that is, it would be allowed to convey information that one could then use to perform logical operations. If there is only 1 way a reaction is maintained, and it consumes gliders or other input that are required to arrive at a specific moment otherwise the pattern breaks, then there is no signal, it is a plain reflector, and calling it "dependent" makes clear it needs periodic gliders to survive, and in that sense is closer to a wick than anything else. So I disagree with the "dependent reflector is a signal reflector" statement, there is no signal, so a dependent reflector is a reflector, which itself is self-explanatory, I think, otherwise on can spell it out and say it is a continuous periodic glider stream reflector. I guess there are 1-time reflectors, that gets consumed it the reflection process, they cannot manage continuous stream of gliders, still calling such a stream "signal" is not meaningful if there is no "start" of that stream, and the number of glider is constant throughout the pattern evolution, with gliders and other misc pattern arranged in some closed loop configuration.

[roolif]

IMHO, a reflection is a change of direction of a straight beam, so it's not applicable to Herschels, Bs, etc, only to spaceships traveling in straight paths.

[confocaloid]

I just edited the page. My edit is explained in the edit summary.
https://conwaylife.com/w/index.php?titl ... did=133133

[dvgrn]

IMHO, a reflection is a change of direction of a straight beam, so it's not applicable to Herschels, Bs, etc, only to spaceships traveling in straight paths.

There's a funny little exception to the first part of that statement: the 0-degree dependent reflectors that are showing up now and again. I think these 0-degree things are interesting enough to get their own section with another table in the dependent reflector article, now that it isn't called "dependent reflector loop" any more. Any objections to that?

If there is only 1 way a reaction is maintained, and it consumes gliders or other input that are required to arrive at a specific moment otherwise the pattern breaks, then there is no signal, it is a plain reflector, and calling it "dependent" makes clear it needs periodic gliders to survive, and in that sense is closer to a wick than anything else.

Hopefully no land mines, just some continuing minor differences of opinion.

I think confocaloid's most recent edit is an improvement, for sure. My original revert was in large part an attempt to defend the existing use of the terms "loop" and "signal". I don't have any objection to the current definition of "dependent reflector".

Now, I do still think that these definitive statements like "there is no signal" are missing the point a little bit, as far as the common meaning of "signal" in the context of Life signal circuitry. Signal circuitry is circuitry that conducts signals. Gliders, spaceships, H, B, R, pi, etc., all generally fall into the category of signals that can be processed by signal circuitry.

It seems to me that it would be a lot less confusing if that "signal circuitry" category can be assumed to include strictly periodic streams of gliders going in to a dependent reflector, and coming out the other side. It's a reflector, which is a type of signal circuitry; the reflector is reflecting a glider stream, and gliders are a member of the "signal" category.

So I'm likely to continue to call those glider streams "signal streams" on occasion, and will also call individual gliders "signals" in the general case where it's useful to do that -- in situations where the Traveling Something is sometimes a glider, sometimes a *WSS, Herschel, B, R, pi, etc., and what I'm talking about is the Traveling Something.

I think that's less confusing than the alternative. Do we really need the little footnotes all the time to say "well, it's definitely not a signal just at the moment, so we shouldn't say that, but it could become a signal again once it leaves the dependent reflector as soon as it gets ORed with another signal or something"? Or should we instead just be saying "dependent reflectors can't transmit binary data", and leave it at that?

[Jormungant]

In a perfect world, everything that is a subclass of some superclass should retain its nomenclature to lay bare the inherent hierarchical structure of term to label items; however, some class do arise because they happen to have properties of two classes, and hence are the intersection of such classes, in this case, either one use both names or makes up a new term to be precise moving forward. "signal streams" are indeed signals if there is some rule that allows arbitrarily timed gliders (periodic with random omission, or with random delays between subsequent gliders) to go through the circuitry, but if the rule is such that only 1 precise configuration is allowed, because things break otherwise, calling that a signal is unprecise in my opinion. I mean, with that logic I would said you would need to look up all the reactions used in crawlers, and call such things "dependent circuits", I think that feel wrong because circuits should have logic, logic implies there are "states" on operates on, and it only 1 state is allowed to exist. I think the confusion is that many state exists in the sense many patterns using units can be defined make different patterns, combined in some combinatorial fashion, but they live in parallel universes: if one state cannot be reached from another, an a circuit filled with continuous glider steams if period N reaches N states precisely, and none of this states intersects that another pattern identical but for some random blinker added that in a corner that does nothing, if there is nothing that can make it disappear, then there is no circuit. I would then say that this N-period is a single state for simplicity, and if it is not allowed to deviate, there is no logic/signal/circuitry, it is an oscillator or a gun, and I would never call the glider steam from a periodic gun a "signal steam", it is a glider stream, nothing travelled, N ticks later we are back where we started. But if a glider stream interacts with some signal stream, with actual information/entropy, then the output will become a signal stream.

In any case, this is my opinion, and I can only speak of what is intuitive to me for the nomenclature, since the English language was never designed to describe patterns in cellular automatons in the first place.

[confocaloid]

IMHO, a reflection is a change of direction of a straight beam, so it's not applicable to Herschels, Bs, etc, only to spaceships traveling in straight paths.

There's a funny little exception to the first part of that statement: the 0-degree dependent reflectors that are showing up now and again. I think these 0-degree things are interesting enough to get their own section with another table in the dependent reflector article, now that it isn't called "dependent reflector loop" any more. Any objections to that?

The 0-degree ones are also described as shifters, as far as I know. For example the stable CP 8-lane glider shifter discovered in March.

There are also other interesting "dependent reactions" of the same kind, such as the p53 and the p69 posted earlier in this thread (I don't know who discovered them). Even though those reactions are not as simple as "one input stream and one output stream", they are still dependent on their input streams.

I agree that these reaction things are interesting, regardless of the terminology.

It seems to me that it would be a lot less confusing if that "signal circuitry" category can be assumed to include strictly periodic streams of gliders going in to a dependent reflector, and coming out the other side. It's a reflector, which is a type of signal circuitry; the reflector is reflecting a glider stream, and gliders are a member of the "signal" category.

A reflector is a stable or oscillating pattern that can reflect some specific type of spaceship without suffering permanent damage. The definition does not demand ability to reflect signals. Even if a device can only redirect a strictly periodic spaceship stream of a given type and period, the device is still a reflector.

In other words I think "a reflector is a type of signal circuitry" is an error. Some (but not all) reflectors can be viewed as signal-processing devices.

When part of an engineered pattern is a "polygonal chain" arrangement of strictly periodic glider streams and dependent reflectors redirecting those streams into each other, then that part of the engineered pattern is not an example of signal circuitry, and there are no signals in it.

[dvgrn]

In other words I think "a reflector is a type of signal circuitry" is an error. Some (but not all) reflectors can be viewed as signal-processing devices.

When part of an engineered pattern is a "polygonal chain" arrangement of strictly periodic glider streams and dependent reflectors redirecting those streams into each other, then that part of the engineered pattern is not an example of signal circuitry, and there are no signals in it.

You still seem to be determined to always use the word "signal" as if it means something like "binary ON-or-OFF compatible signal". The LifeWiki definition of "signal" doesn't say anything even vaguely like that, and never has.

The component patterns in the "Signal-Circuitry" folder in Golly have been used, over the last two decades or so, to build ten thousand pieces of circuitry. Of the gliders in these patterns, hundreds or maybe thousands of them can't possibly act as memory loops or support gaps in the periodic signal streams that they travel in. I have to say "signal" in that last sentence, because the signals are not always gliders -- often it's *WSSes or Herschels or whatever, and often the signals are converted from one form to another at various points.

I have always been able to refer to reflectors and converters and splitters, all of those mechanisms, as "signal circuitry" without any fear that anyone will try to say that's an incorrect label. Until the moment of fuzzy thinking quoted at the top of this thread, it has never even occurred to me that it might be an inappropriate label -- and I still don't think it is. I'd say that an independent reflector is absolutely and always a piece of signal circuitry. In the case of glider reflectors, there are gliders going through them, and those gliders should always be able to be referred to as "signals" -- pieces of information moving through the Life universe.

I'd argue (and, well, yes, I already have, several times!) that dependent reflectors can be safely referred to as "signal circuitry" also, because signal streams clearly go into them, and clearly come out of them. It doesn't matter in the slightest whether binary data can travel through the darn things -- they're still a long-accepted part of a category called "signal circuitry", along with every other type of periodic elementary conduit.

I'm still finding it very odd that there's suddenly this obsessive focus on forbidding the use of the word "signal" for any glider that can't be removed from its stream without damaging the stability of the overall pattern. That information-theory binary OFF/ON encoding meaning of the word "signal" just plain doesn't seem particularly relevant here.

[confocaloid]

I have to say "signal" in that last sentence, because the signals are not always gliders -- often it's *WSSes or Herschels or whatever, and often the signals are converted from one form to another at various points.

Those "gliders or *WSSes or Herschels or whatever" are commonly described as active objects.

In the case of glider reflectors, there are gliders going through them, and those gliders should always be able to be referred to as "signals" -- pieces of information moving through the Life universe.

Again, on the basic level, those gliders are active objects. There are gliders going in, and there are gliders going out. A reflected glider never goes "through" a reflector -- it always becomes some active reaction or merges with an active reaction, and then the reaction spits another glider.

A glider loop with Snarks/rectifiers/bumpers/dependent reflectors/whatever is not a piece of signal circuitry. It is a piece of stationary pattern engineering. A glider loop is an oscillator.

Using the word "signal" with meaning "active object" is bad terminology.

[dvgrn]

Those "gliders or *WSSes or Herschels or whatever" are commonly described as active objects.

They're also, more commonly I think, described as "signals" -- a shorter and less awkward term. I don't think it has ever even occurred to me to use "active object" when I mean "signal".

I'm not sure that this works as an appeal to authority, since I helped write it (though I didn't actually write much of Chapter 4): the Life textbook says on page 100,

An object that moves through a non-empty pattern ... is called a signal.

Notice the emphasis on how it's moving, and not on how it can be either present or absent to carry one bit of information. That's not a requirement.

Footnote at the bottom of the page:

Objects that move through a vacuum (such as gliders) are also sometimes called signals, since they can be thought of as carrying information between two locations.

Nowhere in the section is there any of this unnecessary conflation of the key idea of signals as moving information, with the ideas of "must be able to be present or absent without damaging the pattern" or "must be able to transmit an arbitrary stream of binary data". Your belief that these additional meanings have to be part of any valid definition of "signal" seems to be the root of this disagreement. Ultimately I think you're making things a lot more complicated than they need to be.

Now, this section of Chapter 4 also does not particularly support my contention that dependent reflectors should be considered to be signal circuitry, along with all independent reflectors, switches, splitters, converters and everything else. Also, the great majority of the examples given are in fact potential signals in your bit-carrying sense. This is not surprising, since Chapter 4 is still fairly early on and we're keeping things simple -- no need for four pages of confusing footnotes on information theory.

However, the p5 lightspeed signal on page 101 certainly seems to be an example of something that's definitely called a "signal" in the book, and nobody has objected to that so far. Those p5 signals don't seem to be "active objects", exactly, so that's not a good replacement term. You certainly couldn't remove one of them at any point. You couldn't encode an arbitrary stream of bits using them. That absolutely does not matter. They're p5 lightspeed signals, and that's a nice short clear way to describe them.

A reflected glider never goes "through" a reflector -- it always becomes some active reaction or merges with an active reaction, and then the reaction spits [out] another glider.

A glider loop with Snarks/rectifiers/bumpers/dependent reflectors/whatever is not a piece of signal circuitry. It is a piece of stationary pattern engineering. A glider loop is an oscillator.

Using the word "signal" with meaning "active object" is bad terminology.

This is clearly still the point where we disagree, no matter how many times you re-state all of this.

I do absolutely agree that a reflected glider never goes through a reflector. That's precisely why it's so convenient to say "signal", because signals do go through independent reflectors perfectly well, and signal streams go through dependent reflectors. The signals go in as gliders, and they come out as gliders, but in between they're still (period-N) signals. It's easy and convenient to describe things in this way. I can expect people who work with signal circuitry to understand exactly what I mean without objecting that it is somehow "confusing" or "incorrect" or "bad terminology".

In the dependent reflector case, it's not necessary to care what the specific one-to-one correspondence is between inputs and outputs. A correspondence could theoretically be defined, but in practice, which specific correspondence is chosen doesn't matter to the functioning of the mechanism, so people building complex signal circuitry won't usually waste any brain cells on the question. And they shouldn't have to, just to be able to use the term signal "correctly".

TL;DR:
Your insistence on a narrow interpretation of "signal" to mean only "bit-carrying signal" seems to be getting you tied up in linguistic knots. You won't find a lot of mentions of "active object" in the Life textbook in contexts where "signal" is meant, because that's unnecessarily awkward. I think that your overly narrow definition of "signal" would create a whole lot more confusion, compared to the long-standing common usage that you're objecting to.

I think we should continue with the convention of considering "signal circuitry" to be everything that carries and processes (what I'm calling) signals.

[confocaloid]

This is an incomplete reply, but it is a reply.

I hope the following parts make it sufficiently clear that you're replying specifically to me and addressing me:

... Your belief ...
... Your insistence ...

You still seem to be determined [...]
...
I'd argue (and, well, yes, I already have, several times!) [...]
...

Please let other members to participate in this forum discussion, so that they could give their opinions on the matter. As it stands, I think you are ignoring feedback from other members who already posted here, and instead are trying to promote your point of view.
In this forum thread, there were already members who participated in this discussion, and did not express agreement with your point of view.
In this forum thread, there were already members who explicitly expressed disagreement with you in some way or another.
Please do not ignore feedback.

[dvgrn]

Please do not ignore feedback.

Feedback from other people, here on the forums, is exactly what I am hoping for at this point. We have probably both explained our positions as clearly as we are capable of.

[Jormungant]

Well, I got to say that I agree with confocaloid, at least the definition of signal makes me think of an "event" or "message passing" in computer science, with one infamous "Ctrl-c signal" being the your friendly equivalent to alt-f4, which is also known on the linux side as "sigkill". In that sense, a signal should be reacted to, or at least imply some message is being interpreted and cause further downstream events/signals, but based on some logic.

That being said, if you ask me to find a jargon term for "an object that moves through a non-empty pattern", I have nothing obvious that pops in my mind at all, it is a very vague concept that has no clear equivalency in our physical world. So I guess one could call that a signal, it is better than "linkable pattern with deterministic evolution in bounded structure", for instance. From this point of view, I guess there is nothing wrong with calling this a signal, people could probably have equivalent but dissimilar criticisms if we changed that to "ripple" or "wave", so "signal" is the current placeholder, as imperfect as it may be.

[glider_rider]

I generally think they should be considered signals, or at least that there's nothing wrong with calling them that. They don't carry as much information as a stream of bit-carrying signals, and said information is generally far less useful, but they still carry information. If you remove a glider from a reflector loop, assuming the reflectors don't interfere with one another when they explode, the pattern stops working with the flow of the gliders rather than all at once.

I mean, you could even have more complicated cases. For example, gliders fired into an independent reflector modified to also tick down a counter for every missing glider and explodes when it reaches 0. Basically 'a dependent reflector which can tolerate n missing inputs'. This also can't encode a stream of bit-carrying signals, but it definitely can transmit information, and in fact more information than a standard dependent reflector (some predefined finite number of missing glider positions). (Of course if you don't view it as a black box it obviously contains bit-carrying signals internally, but it's not inconceivable there could be a reflector with these properties that isn't just a self-sabotaging independent reflector, and that wouldn't have bit-carrying signals anywhere. This is why I like viewing things as black boxes.)

[dvgrn]

Thanks for the input, everyone. I don't think enough people have responded yet to count as any kind of accurate survey of the community, but it has at least been shown that neither confocaloid nor I are alone in our opinions.

Since we're kind of stuck on a To Edit Or Not To Edit problem until this is resolved, it'd maybe be nice if about a dozen more new people would chime in here -- maybe just with brief answers to the following questions:

1) is it okay to describe closed period-N Snark loops as "signal circuitry"?
2) is it okay to describe closed period-N dependent reflector loops as "signal circuitry"?
3) is it okay to rename the p49 glider shuttle article to "p49 glider loop", with a redirect?
4) is it okay to rename the p26 glider shuttle, p31 glider shuttle, and p59 glider shuttle to call them all "loops" instead, with a redirect? EDIT: Fixed a duplication. I would think these three would best be called "p{N} dependent glider loops" or possibly "p{N} dependent reflector loops".
5) is it okay to rename the p71 glider shuttle to "p71 dependent reflector loop" while we're at it, with a redirect? (Not sure if anyone might think that's different from the others, being 180 degrees and thus maybe a bit more shuttle-like.)

[confocaloid]

However, the p5 lightspeed signal on page 101 certainly seems to be an example of something that's definitely called a "signal" in the book, and nobody has objected to that so far. Those p5 signals don't seem to be "active objects", exactly, so that's not a good replacement term.

I want to clarify that I know "signal" is used to refer to deformations that move through a non-empty background pattern. I do not have any complaints to that. The section 4.5.1 is titled "Wires and Signals" and related terms are defined.

The wiki page Signal currently gives a different meaning of "signal" (an information carried by some moving object). In this meaning, there is a useful distinction between the information and the object that carries it.

[johamwit]

1) is it okay to describe closed period-N Snark loops as "signal circuitry"?

YES, they can be used as MEMORY LOOPS

2) is it okay to describe "closed period-N dependent reflector loops" as "signal circuitry"?

NO, they are perfectly described as closed period-N dependent reflector loops. It's a type of oscillator in which the rotor includes glider streams.

3) is it okay to rename the p49 glider shuttle article to "p49 glider loop", with a redirect?

YES, it's a MEMORY LOOP storing 4 bits. As all MEMORY LOOPS, it can be viewed as an oscillator :
0000 are 4x p7 oscillators, 0001 is a p196, 0101 is a p98 and 1111 is a p49,

4) is it okay to rename the p26 glider shuttle, p31 glider shuttle, and p59 glider shuttle to call them all "loops" instead, with a redirect?

NO, they are just oscillators with a rotor composed partially of glider streams. If you touch the rotor, the oscillator is destroyed. If you want to call them "loops", the word "dependent" should be used.

5) is it okay to rename the p71 glider shuttle to "p71 dependent reflector loop" while we're at it, with a redirect?

YES, it fulfills the criteria used so far. A DR-LOOP, something that visually resembles a "true" glider loop.

[Anivec]

Should this also go on the dependent reflector list?

p47 dependent reflector based off the gun partial, incredibly silly:

Code: Select all

x = 127, y = 127, rule = B3/S23
89bo7bo$88bobo5bobo$89bo7bo3$90bo5bo$89bobo3bobo$83b3o15b3o$89b3o3b3o$
80b2o23b2o$79bobo23bobo$79bo27bo$78b2o27b2o2$85b2o13b2o$85b2o13b2o2$
61b2o15b2o27b2o$62bo16bo27bo$12bo4bo40b2o2bob2o13bobo23bobo$10b3o4b3o
38bobobobo2bo12b2o23b2o$9bo10bo39bo2bo2b2o21b3o3b3o$9b2o8b2o38b2o22b3o
15b3o$89bobo3bobo$7bo14bo67bo5bo$7bo14bo$7bo6b2o6bo30b2o6b2o13b2o$14b
2o37bo6bo2bo11bo2bo10bo7bo$10b2o6b2o30b2obo6bo3bo11bob2o8bobo5bobo$bo
26bo22bob2o6bo2bo12b2o10bo7bo$obo4bo14bo4bobo21bo9bo2bo$bo4b2o14b2o4bo
23b5o5b2o$7bo14bo26b3o4bo$48bo3bobo22b2o$49b2obob2o20bo2bo$50bobo21b2o
b2o4b2o$7bo14bo26bo2b4o6b2o10b2ob2o2bo2bo$bo4b2o14b2o4bo20bobo3bo6b2o
12bo4b3o$obo4bo14bo4bobo20bob3o21b2o$bo26bo23bo23b2o3b2obob2o$10b2o6b
2o19b2o9bobo4b3o16b2o4bob2obo$14b2o23bo10b2o5bo24bo$7bo6b2o6bo17bo17bo
22b2o$7bo14bo12b2o2b2o$7bo14bo12bobo26bo$37bob4o21bobo$9b2o8b2o15b2obo
2bo21b2o$9bo10bo72bo$10b3o4b3o69b2obobo$12bo4bo10b2o3b2o50b2obo2b2obo
3bo$26b2obo3bob2o2b2o44bo3bo4bob3o$26b3o7bo2b2o29bo15b3ob4obo$34b3o32b
2o17bobo4bob4o$69bobo16bobob2obobo2bo$89b2obo2bo$94b2o$52bobo$27b2o23b
2o30b3o19b2o$53bo32bo17bo2bo$85bo12b2o4b3o$20b2o77bo9bo$21bo23bo43b2o
4b2ob2o6b4o$19bo23bobo43b2o5bobo6bo$19b2o8bo14b2o35b2o14bo8b2o$21bo6bo
bo5b2o43bobo23bo$17b4o6b2ob2o4b2o43bo23bo$17bo9bo77b2o$20b3o4b2o12bo$
19bo2bo17bo32bo$19b2o19b3o30b2o23b2o$72bobo$31b2o$31bo2bob2o$26bo2bobo
b2obobo16bobo$26b4obo4bobo17b2o32b3o$31bob4ob3o15bo29b2o2bo7b3o$28b3ob
o4bo3bo44b2o2b2obo3bob2o$28bo3bob2o2bob2o50b2o3b2o10bo4bo$32bobob2o69b
3o4b3o$33bo72bo10bo$61b2o21bo2bob2o15b2o8b2o$60bobo21b4obo$62bo26bobo
12bo14bo$86b2o2b2o12bo14bo$44b2o22bo17bo17bo6b2o6bo$44bo24bo5b2o10bo
23b2o$39bob2obo4b2o16b3o4bobo9b2o19b2o6b2o$39b2obob2o3b2o23bo23bo26bo$
49b2o21b3obo20bobo4bo14bo4bobo$43b3o4bo12b2o6bo3bobo20bo4b2o14b2o4bo$
42bo2bo2b2ob2o10b2o6b4o2bo26bo14bo$42b2o4b2ob2o21bobo$47bo2bo20b2obob
2o$48b2o22bobo3bo$70bo4b3o26bo14bo$63b2o5b5o23bo4b2o14b2o4bo$62bo2bo9b
o21bobo4bo14bo4bobo$29bo7bo10b2o12bo2bo6b2obo22bo26bo$28bobo5bobo8b2ob
o11bo3bo6bob2o30b2o6b2o$29bo7bo10bo2bo11bo2bo6bo37b2o$49b2o13b2o6b2o
30bo6b2o6bo$104bo14bo$30bo5bo67bo14bo$29bobo3bobo$23b3o15b3o22b2o38b2o
8b2o$29b3o3b3o21b2o2bo2bo39bo10bo$20b2o23b2o12bo2bobobobo38b3o4b3o$19b
obo23bobo13b2obo2b2o40bo4bo$19bo27bo16bo$18b2o27b2o15b2o2$25b2o13b2o$
25b2o13b2o2$18b2o27b2o$19bo27bo$19bobo23bobo$20b2o23b2o$29b3o3b3o$23b
3o15b3o$29bobo3bobo$30bo5bo3$29bo7bo$28bobo5bobo$29bo7bo!

Edit: 180 degree dependent if that's your thing:

Code: Select all

x = 137, y = 148, rule = B3/S23
12bo4bo$10b3o4b3o$9bo10bo$9b2o8b2o3$6b3o12b3o$14b2o$14b2o2$bo26bo$obo
6bobo6bobo6bobo$bo7bo2bo4bo2bo7bo$9bobo6bobo4$9bobo6bobo$bo7bo2bo4bo2b
o7bo$obo6bobo6bobo6bobo$bo26bo$39b2o$14b2o23bo$14b2o24bo$6b3o12b3o11b
2o2b2o$31bo3bobo$37bob4o$9b2o8b2o15b2obo2bo$9bo10bo$10b3o4b3o$12bo4bo
12bobo$30b3o6b2o$31bo7b2o7$40bobo$29b2o10b2o$20b2o8bo10bo$21bo6bobo$
19bo6b2o$19b2o4b2o$21bo5bo8b2o$17b4o5bo2bo6b2o$17bo7b2ob3o9bo$20b3o5bo
2bo7bobo$19bo2bo2b2o12bobo2b2o$19b2o17b2ob4o8bo$46bo7bo$31b2o10b4o5b3o
$31bo2bob2o5bo$26bo2bobob2obobo5b3o$26b4obo4bobo2bo4bo$31bob4ob4o$28b
3obo4bo$27bo2bobobo2bob3o$28b2o3b2o3bobo7bo$42bo4b2o$41b2o4bobo$64bo$
65b2o$64b2o7$59b2o$59bobo$59bo$77bo$75bobo$76b2o7$71b2o$70b2o$72bo$87b
obo4b2o$88b2o4bo$88bo7bobo3b2o3b2o$95b3obo2bobobo2bo$99bo4bob3o$95b4ob
4obo$90bo4bo2bobo4bob4o$90b3o5bobob2obobo2bo$93bo5b2obo2bo$82b3o5b4o
10b2o$82bo7bo$83bo8b4ob2o17b2o$91b2o2bobo12b2o2bo2bo$95bobo7bo2bo5b3o$
96bo9b3ob2o7bo$99b2o6bo2bo5b4o$99b2o8bo5bo$110b2o4b2o$109b2o6bo$106bob
o6bo$95bo10bo8b2o$94b2o10b2o$94bobo7$96b2o7bo$96b2o6b3o$104bobo12bo4bo
$117b3o4b3o$116bo10bo$94bo2bob2o15b2o8b2o$94b4obo$99bobo3bo$96b2o2b2o
11b3o12b3o$96bo24b2o$97bo23b2o$96b2o$108bo26bo$107bobo6bobo6bobo6bobo$
108bo7bo2bo4bo2bo7bo$116bobo6bobo4$116bobo6bobo$108bo7bo2bo4bo2bo7bo$
107bobo6bobo6bobo6bobo$108bo26bo2$121b2o$121b2o$113b3o12b3o3$116b2o8b
2o$116bo10bo$117b3o4b3o$119bo4bo!

[dvgrn]

4) is it okay to rename the p26 glider shuttle, p31 glider shuttle, and p59 glider shuttle to call them all "loops" instead, with a redirect?

NO, they are just oscillators and guns with a rotor composed partially of glider streams. If you touch the rotor, the oscillator is destroyed.
Hehe... that p59 is a special case. BTW, it's best described as a p59 gun...

Just to check... are you actually answering the question about the p59 glider shuttle here (the pattern in the infobox)? Or are you instead looking at the reduced p59 gun that shows up in the main article, as an example use of the dependent reflector?

I'm asking partly because, given the "NO" to #4, your answer to #5 seems a little surprising -- since the p71 is a dependent reflector loop quite similar to the others, except for being made up of two 180-degree reflectors instead of four 90-degree ones. Did you mean to give it a YES where the other three all get a NO?

[johamwit]

I apologize, of course I was looking at the main image, the p59 gun, instead of the pattern in the infobox, as made clear by the rest of the paragraph. So, my answer to the three examples given at 4) is NO.

And yes, as I look now at that p59gun, the three instances of the same Dependent Reflector are there, connected by small glider paths.

I didn't pay enough attention, sorry. For the sake of clarity, I'll edit out the p59g rant

[Jormungant]

I think I would like to know what are the advantage using the term "signal circuitry" over plain "circuitry". I think the ambiguity of "circuitry" seems to be more appropriate for the grandly overreaching term you mean to use (I think). One could think of logic circuits, which is to only thing one would think it refers to if you restrain it to "signal circuitry", but also of racing tracks, where our little glider or xwss are bouncing around, being rephased, or sometime even involving some stargate. In the later case, the only thing that matters is that there is a deterministic input -> output track, how many gliders with no entropy, or signal if it abides by whatever your personal preferred definition of "signal" might be, does not matter because it is not explicitly specified.

I mean, I can support the idea that one defines the term "signal", with is great for communication and consistency; but in practice, if the "signal" happens to be a glider, I would just say "the glider" to be more precise, and if I cant, I would use "the input", "the output", "the result", I would personally avoid using the term just for clarity sake.

Otherwise, I would also agree with johamwit on the 5 questions asked. The term shuttle is more precise than loop, to me it means "loop with dependent reflectors", but that might be clear to me just because how people named such patterns in the past.

[dvgrn]

Otherwise, I would also agree with johamwit on the 5 questions asked. The term shuttle is more precise than loop, to me it means "loop with dependent reflectors", but that might be clear to me just because how people named such patterns in the past.

As far as I can tell, we have about forty years of Life-related discussions where "shuttle" always meant one active region, moving back and forth... followed by just a few occasional uses in the less precise sense, starting in the last decade. See the Life Lexicon definition. The term started with the queen bee shuttle and twin bees shuttle and consistently meant the same thing for four decades.

On Discord, hotdogPi pointed out that the 2005 p72 quasi-shuttle had to be called "quasi", specifically because "although this looks at first sight like a shuttle, it isn't really, because each side of the oscillator uses a different hassling reaction".

Long story short, it seems to me that if we think we need to find a word that means "loop with dependent reflectors", it would be a mistake to use either "shuttle" or "relay". Those already have long-standing precise meanings that don't really match what we need.

"Relay"
I could be wrong about "relay". The Life Lexicon and LifeWiki both say "Any oscillator in which spaceships (typically gliders) travel in a loop". So it's clear that "relay" is different from "shuttle", because the object being relayed has to be a spaceship, where in a shuttle it often is not a spaceship. But so far I can't find any examples where a relay has multiple spaceships in its loop -- it was always just one, like in the glider relays in the various Primer patterns.

So all prior usage that I've found so far seems to imply that "relay" is pretty much a synonym for "spaceship shuttle" -- or maybe I should say for "spaceship quasi-shuttle". If we suddenly restrict "relay" to mean only "loop with dependent reflectors", that would be ... a little bit odd.

[confocaloid]

1) is it okay to describe closed period-N Snark loops as "signal circuitry"?

A period-N Snark loop is best described as an oscillator. The concepts of an oscillator and a reflector are simple (Chapter 3, section 3.5 in the textbook) in comparison to concepts that arise in logic/circuitry/signal discussions.
One could describe a period-N Snark loop as "signal circuitry", should they choose to view it that way. But it does not do much "signal processing" by itself.

2) is it okay to describe closed period-N dependent reflector loops as "signal circuitry"?

No, those are oscillators made out of dependent reflectors and segments of uniformly-spaced spaceship streams. Unlike a Snark loop, a dependent reflector loop can be "filled" in only one way.

3) is it okay to rename the p49 glider shuttle article to "p49 glider loop", with a redirect?

From the viewpoint of a wiki editor, my preference is to keep the existing name. The article is named "p49 glider shuttle" for many years, and there were no objections/discussions of the name until now (as far as I can tell).

4) is it okay to rename the p26 glider shuttle, p31 glider shuttle, and p59 glider shuttle to call them all "loops" instead, with a redirect? EDIT: Fixed a duplication. I would think these three would best be called "p{N} dependent glider loops" or possibly "p{N} dependent reflector loops".
5) is it okay to rename the p71 glider shuttle to "p71 dependent reflector loop" while we're at it, with a redirect? (Not sure if anyone might think that's different from the others, being 180 degrees and thus maybe a bit more shuttle-like.)

From the viewpoint of a wiki editor, my preference is to keep the existing names for the first two articles. Those were named many years ago, and their names appear in existing discussions.
The page p59 glider shuttle and p71 glider shuttle are created recently. For these two pages, I don't have a preference one way or another.

[Anivec]

1) is it okay to describe closed period-N Snark loops as "signal circuitry"?

No, simply because there is no input or output from the loop to except, if it was actually connected to circuitry then yes, it would.

YES, they can be used as MEMORY LOOPS

Counter argument, it’s not a memory loop because there is no way to read it or write in it without involvement of external circuitry, meaning the “memory” isn’t actually memory until it’s connected externally, thus it’s not a piece of signal circuitry by itself

2) is it okay to describe "closed period-N dependent reflector loops" as "signal circuitry"?

Also no, this time because it’s completely dependent. You can’t use it for much signal circuitry.

3) is it okay to rename the p49 glider shuttle article to "p49 glider loop", with a redirect?

Yes, I always thought a shuttle was a glider going 180 degrees without any 90 degree turns, I might be wrong though. As long as you call it a p7 dependent loop I’ll be fine

4) is it okay to rename the p26 glider shuttle, p31 glider shuttle, and p59 glider shuttle to call them all "loops" instead, with a redirect?

Yes, but only if you call them dependent loops for a good reason

5) is it okay to rename the p71 glider shuttle to "p71 dependent reflector loop" while we're at it, with a redirect?

Yeah, sure, you could call it a dependent shuttle though.
I don’t have much concern for 4 and 5, mainly because I wanted to tell this one person that memory loops are not the same thing as a normal glider loop

[Jormungant]

I also do not have a strong opinion on 4 and 5, I even forgot that twin-bee "shuttle" thing, that very concept just got overwritten by "it can become a gun". I am not sure what is your precise definition for shuttle is, but if it only involve some symmetry, as it moves "back and forth", not sure why rotation symmetry is not allowed in shuttles, even if it has odd number of gliders bouncing around. Well, if they have to be renamed, I think I prefer "relay" over "loop".

[confocaloid]

You wouldn't normally have a relay with multiple simultaneous signals in it, as far as I can tell.

But so far I can't find any examples where a relay has multiple spaceships in its loop -- it was always just one, like in the glider relays in the various Primer patterns.

By definition, the p43 Snark loop is a relay. It is an oscillator in which gliders travel in a loop. So it is an example where a relay has multiple spaceships.

By the existing definition, a "dependent reflector loop" is not a relay. However, the reason is different. In a "dependent reflector loop", the individual spaceships only appear to travel in a loop, but don't really travel in a loop all the way round. So these "dependent reflector loops" are something that visually resembles a relay, but isn't a relay.

The definition of "relay" says:

:relay Any oscillator in which spaceships (typically gliders) travel in a loop. The simplest example is the p60 one shown below using two pentadecathlons. Pulling the pentadecathlons further apart allows any period of the form 60+120n to be achieved - this is the simplest proof of the existence of oscillators of arbitrarily large period.

...........................O....O..
................OO.......OO.OOOO.OO
.................OO........O....O..
................O..................
..O....O...........................
OO.OOOO.OO.........................
..O....O...........................