Life Lexicon update collection thread

For discussion directly related to LifeWiki.
User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Life Lexicon update collection thread

Post by dvgrn » November 25th, 2021, 4:44 pm

Lexicographers Wanted
It's been almost three and a half years now since the Life Lexicon got an update. The longer it goes between updates, the harder it becomes to get caught up... which is exactly why it once went eleven years between updates.

Three years isn't too bad, yet. We can probably dig up most of the new terminology that needs to be added, by looking through the News archive on the LifeWiki and the Pattern of the Year competitions. But I could still definitely use some help.

"Why bother?"
There is a lot of overlap these days between the Life Lexicon and the LifeWiki, but I think it's still worth keeping the Lexicon up-to-date within half a decade or so. In good part this is because the Lexicon travels with Golly. There are all kinds of useful small patterns that I get out of the Lexicon when I need them for a construction project. But nowadays, if I want to go and find a Bandersnatch that way, I'm out of luck.

The latest version of the Lexicon is kept in this Git repository. But since nobody has ever even once contributed anything to that repository... my sense is that a more convenient collection mechanism is needed for new definitions. So let's try a forum thread!

Ground Rules (Sort Of)
Similar to contributions to Golly's pattern collection, the more "finished" a definition is when it's posted here, the more likely that it will actually make it into the Lexicon. "We need a definition for X" might end up getting cheerfully ignored, but if someone makes the effort to produce an actual Lexicon-compatible short definition, I'll be much better motivated by not wanting to let all that work go to waste.

Head Lexicographer Especially Wanted
If someone would be interested in taking over and leading this next Lexicon update project, that would be great too -- as long as it's someone who can pay close attention to all the fiddly typographical, grammatical, and historical details, so that the update will actually be an improvement! David Bell and I have gotten through a couple of rounds of Lexicon updates now, but it's (fair warning!) a whole lot of work. Even so, it might be time for it to be somebody else's turn.

The Plan, Such As It Is
The current to-do list is here. I'll start by adding more terms like "Bandersnatch" to that list, I guess, and then remove them from the list as the final definitions get added to the individual files in the repository.

hotdogPi
Posts: 1586
Joined: August 12th, 2020, 8:22 pm

Re: Life Lexicon update collection thread

Post by hotdogPi » November 25th, 2021, 8:08 pm

Patterns can probably be taken from the "patterns found in 2019/2020/2021" thread, with the exception of the U-turner and Octomino II which aren't there because they're not discoveries but can still be in the Lexicon. "Iwona active region" also seems to be the standard name for this object:

Code: Select all

x = 5, y = 2, rule = B3/S23
ob3o$2o!
In addition, "index fossil" is probably something that you're the best at describing.
User:HotdogPi/My discoveries

Periods discovered: 5-16,⑱,⑳G,㉑G,㉒㉔㉕,㉗-㉛,㉜SG,㉞㉟㊱㊳㊵㊷㊹㊺㊽㊿,54G,55G,56,57G,60,62-66,68,70,73,74S,75,76S,80,84,88,90,96
100,02S,06,08,10,12,14G,16,17G,20,26G,28,38,47,48,54,56,72,74,80,92,96S
217,486,576

S: SKOP
G: gun

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » August 11th, 2022, 4:36 pm

Just looked at this again, and it had been a couple of years since I had even updated the to-do list for this project.

I'd like to get a new Lexicon release put together, not for the immediately upcoming Golly 4.2 release, but for Golly 4.3 whenever that happens.

"Ready-for-editing" candidate definitions
As usual, I'll be hoping that people can contribute complete definitions here that they think could be added to the Lexicon. Lists of terms that need to be defined, but aren't yet ... are not so terribly helpful, and therefore might not get much attention. I always end up doing lots of editing anyway, but it really speeds things up when I'm not starting from a blank page.

I think quite a lot of definitions can be copied and abridged straight from LifeWiki articles. However, there will clearly be a lot of arbitrary judgment calls this time around: the Lexicon is not the LifeWiki, and not every pattern documented in a LifeWiki article really needs a Lexicon definition.

What's Worth Including?
There are three general categories of new terms that I think should be included:

1) New terminology that probably didn't exist the last time a Lexicon release happened. These entries answer questions like "what does SKOP mean?" or "what is a DBCA?"
2) New large engineered objects that have names and are referred to regularly, but can't be displayed as an ASCII pattern. A good starting point is the POTY nominations list, so I've added 2020 and 2021 POTY nominees to the to-do list. Not all of them need Lexicon entries, of course, but some may inspire several entries. Examples: "DOGun SaGaQR", "SSOL", "Remini", "QuickSilver Demonoid". Conversely, if something didn't get a POTY nomination, it might not be getting discussed often enough to need a Lexicon entry (e.g., maybe a LifeWiki article is good enough).
3) New named objects that fit inside 64x64, that are likely to be used as components of larger constructions -- e.g., "Bandersnatch", "meatball", "speed tunnel", "Charity's p16". It's nice to have these kinds of things within reach of easy copy/pasting from within Golly, so it might be good to err slightly on the side of including more of these rather than less -- even though more complete information will almost inevitably be available in LifeWiki article form.

New HTML display functionality?
I haven't looked into this yet, but with the upcoming update in Golly 4.2 to use wxWebRequest instead of wxHTML (which was done to support https:// links) maybe there's something that could be done to improve the presentation of the Lexicon a bit further, especially getting a Find function working across all Lexicon text. Might be a forlorn hope, but once that change is in, it seems worth having a fresh look.

Book
Posts: 385
Joined: August 28th, 2021, 2:38 pm
Location: California
Contact:

Re: Life Lexicon update collection thread

Post by Book » August 11th, 2022, 4:59 pm

Lexicon appears to be used as an authoritative source for a lot of wiki info. My concern is how we avoid getting into a loop (the source for the wiki is lex which was sourced from the wiki which...).
Phil Bookman

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » August 11th, 2022, 5:11 pm

Book wrote:
August 11th, 2022, 4:59 pm
Lexicon appears to be used as an authoritative source for a lot of wiki info. My concern is how we avoid getting into a loop (the source for the wiki is lex which was sourced from the wiki which...).
It's probably not a huge problem in practice. The Life Lexicon had a 2006 edition, which was a major source of information for Nathaniel's original set of articles when the LifeWiki was created in 2009. That's bound to give the impression that the Life Lexicon is a source for LifeWiki information.

However, for topics that didn't exist in 2006, the LifeWiki article is very very likely to have come into existence long before the Life Lexicon entry. We could probably even make that be standard practice: anyone who wants a given Life Lexicon entry to exist, should start by making a LifeWiki article of the same name, or a redirect where appropriate.

The authoritative source will already have been documented in the LifeWiki article, and that reference information does not have to be copied into the Lexicon entry -- we can do short definitions in the Lexicon, with the understanding that more background, links to sources, etc., will be available in the corresponding LifeWiki article. I'll add something early on in the Lexicon Introduction page to point out that assumption.

hotdogPi
Posts: 1586
Joined: August 12th, 2020, 8:22 pm

Re: Life Lexicon update collection thread

Post by hotdogPi » August 18th, 2022, 9:30 am

Null reaction: A reaction between two or more objects that has no net effect, as if the objects didn't react at all.

Index fossil: A rare object, or more rarely a methuselah or diehard, that almost always forms from the same seed of just a few small common objects surrounded by empty space. Given the object, you can typically tell how it formed because that's pretty much the only way it forms.

The term "fallback" is described quite well on the wiki.

Leading edge: The configuration of the front row of an active region. Usually, it is described as the number of cells, such as a two-cell leading edge, but nonconsecutive cells may require a more detailed description such as obo.

Partial result (already exists): add a paragraph saying "For high-period oscillators, partial results can also be the result of objects reappearing in the correct position but with perturbations that can't be made with sparks or catalysts and/or junk that can't be deleted."

apgcode: A way of uniquely identifying objects using short strings of text. While the code is not very human-readable, the prefix tells you what type of object it is.

Hensel notation: The primary way of notating range-1 Moore neighborhood isotropic non-totalistic rules.

Shield pentaplet: b3o$o3bo!
U-turner: 2o$ob2o$b3o$2b2o!, or anything in the same sequence
R-turner: bo$obo$o2bo$3bo$2b2o!, or anything in the same sequence
Iwona active region: ob3o$2o!, or anything in the same sequence
Sandwich: this sequence: b3o$o3bo$2o$2b3o!
Toast: this sequence: 3bo$b3o$o3bo!
Octomino II: bo$3o$ob2o$3bo!
Octomino II to glider: 3o$b3o$5b3o$5bo2bo$4bo$3bo!
User:HotdogPi/My discoveries

Periods discovered: 5-16,⑱,⑳G,㉑G,㉒㉔㉕,㉗-㉛,㉜SG,㉞㉟㊱㊳㊵㊷㊹㊺㊽㊿,54G,55G,56,57G,60,62-66,68,70,73,74S,75,76S,80,84,88,90,96
100,02S,06,08,10,12,14G,16,17G,20,26G,28,38,47,48,54,56,72,74,80,92,96S
217,486,576

S: SKOP
G: gun

Book
Posts: 385
Joined: August 28th, 2021, 2:38 pm
Location: California
Contact:

Re: Life Lexicon update collection thread

Post by Book » August 18th, 2022, 3:55 pm

How about a def for "active region"?
Phil Bookman

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » March 11th, 2023, 4:47 pm

When discussing conduits/hasslers, I think "active reaction" is a better way to convey the intended meaning.

Compare (the CGoLM&C book, 7.1.1 Conduit Terminology):
"...as the active reaction passes through the conduit.",
"Like an eater, a catalyst is not permanently affected by a passing active reaction"

My two issues with "active region" is that
Book wrote:
August 18th, 2022, 3:55 pm
How about a def for "active region"?
Even though that word combination is in fact commonly used (e.g. on this forum), there may be a different better clearer wording which is also commonly used.

AFAIK "methuselah" is also sometimes used to mean the same thing (example) -- any chaotic mess that is useful/looks potentially useful for conduits/hasslers.
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » April 1st, 2023, 12:13 pm

I've been strenuously avoiding the next Life Lexicon update because it's too darn much work. But it occurred to me recently that there's a new option in the world that wasn't available last time the Lexicon needed an update: I can ask an AI to write Life Lexicon entries for me!

I tested ChatGPT with a sample question:
dvgrn> What is the Life Lexicon?

ChatGPT The Life Lexicon is a comprehensive list of patterns, structures, and terms used in the study of cellular automata, specifically Conway's Game of Life. It was compiled by Stephen Silver, a mathematician and computer scientist, and contains over 1300 entries that describe various types of objects and behaviors that can occur in the Game of Life.

The Life Lexicon serves as a reference for researchers and enthusiasts of cellular automata and provides a standardized vocabulary for discussing and describing patterns and structures in the Game of Life. Entries in the lexicon include descriptions of oscillators, spaceships, still lifes, and other objects, as well as terms for different types of interactions between these objects.

The Life Lexicon is considered an important resource for those interested in Conway's Game of Life and has been referenced in numerous publications on the subject.
That seemed pretty darn good! So once I knew that ChatGPT was trustworthy, I got rolling and started collecting definitions.

At this rate I'll be done with the next Lexicon update in almost no time at all:
dvgrn> Can you write a Life Lexicon style definition for "minstrel"?

ChatGPT> Certainly! Here's a Life Lexicon style definition for "minstrel":

A "minstrel" is a period 4 oscillator that was discovered by Achim Flammenkamp in 1994. It is composed of two copies of the "pentadecathlon" oscillator, which are connected by a "block" and two "tubs" in a specific arrangement. The two pentadecathlons oscillate at right angles to each other, with a phase difference of 90 degrees. The minstrel has a large number of active cells, making it useful for constructing larger patterns.
It must be true, because if it wasn't true, nobody could possibly have thought of inventing it... nobody besides ChatGPT, anyway.
dvgrn> Can you write Life Lexicon style entries for "upconverter" and "downconverter"?

ChatGPT> Sure, here are the Life Lexicon style entries for "upconverter" and "downconverter":

"Upconverter": An "upconverter" is a type of spaceship that is capable of converting a glider into a higher frequency object. The simplest upconverter is the "HFx86" which was discovered by Nicolay Beluchenko in 2009. It works by colliding with a glider and producing a Herschel along with other debris. The Herschel can then be used to synthesize higher frequency objects. Upconverters have been used to construct a wide range of complex patterns in Conway's Game of Life.

"Downconverter": A "downconverter" is a type of spaceship that is capable of converting a higher frequency object into a glider. The simplest downconverter is the "LWSS-to-G" which was discovered by David Buckingham in 1972. It works by colliding with a lightweight spaceship and producing a glider along with other debris. Downconverters have been used to construct a wide range of complex patterns in Conway's Game of Life.
dvgrn> Can you write a Life Lexicon style entry for "doubled-signal toolkit"?

ChatGPT> Certainly! Here's a Life Lexicon style entry for "doubled-signal toolkit":

"Doubled-signal toolkit": The "doubled-signal toolkit" is a collection of stable circuitry in Conway's Game of Life that is capable of transmitting and receiving two signals simultaneously along the same wire. It was first discovered by Matthias Merzenich in 2012, and has since been used in the construction of various complex patterns.

The toolkit includes a "double-barreled shotgun" and a "double-bit reaction" among other components. The double-barreled shotgun is a stable circuit that can emit two gliders at once, while the double-bit reaction is a stable circuit that can be used to create a pair of gliders in response to a single input. The toolkit also includes various other stable circuits that are capable of receiving and transmitting two signals along a wire in various ways.

The doubled-signal toolkit has been used to construct various complex patterns that require the transmission and reception of multiple signals, and has been an important development in the study of cellular automata.
... On that last one, I kept downvoting the answer and telling it what was wrong. Apparently completely at random, it kept picking different names and dates for the origin of the toolkit. After the third revision it stumbled on one of the correct discoverers, and kind of got the first and last paragraphs mostly right, but confidently put the date two years before Herschel conduits were even discovered:
Third Try
ChatGPT> Certainly! Here's the Life Lexicon style definition for "doubled-signal toolkit":

A "doubled-signal toolkit" is a collection of patterns and techniques that can be used to manipulate two signals that are travelling in parallel through a circuit in Conway's Game of Life. The toolkit was first proposed by Dave Greene in 1994 and is based on the idea of using Herschel conduits to double a signal.

The toolkit includes several key components, including a "signal doubler" which can be used to split a signal into two identical copies, and a "signal splitter" which can be used to separate two signals that are travelling in parallel. The toolkit also includes a range of other components such as reflectors, filters, and amplifiers that can be used to manipulate the signals in various ways.

The doubled-signal toolkit has proven to be a powerful tool for constructing complex circuits and has been used to build a wide range of patterns, including adders, multipliers, and memory cells. Its ability to manipulate multiple signals simultaneously has made it particularly useful for constructing large-scale digital circuits in Conway's Game of Life.
Then In the grand Gustavo tradition, maybe we need to use the doubled-signal toolkit to build some of these things that ChatGPT claims the existence of (adders, multipliers, and memory cells)... I suppose that the initial G0 converter in the doubled-signal chain is the "amplifier" -- I didn't know that term.

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » July 16th, 2023, 7:34 pm

Cross-posting here my complaint about the current definition of "track":
confocaloid wrote:
July 16th, 2023, 7:29 pm
(*) I'm using the following definition of "track", taken from an old version of Life Lexicon:
Life Lexicon (2002-05-23) wrote::track A path made out of conduits, often ending where it begins so that the active object is cycled forever, forming an oscillator or a gun.
Unfortunately, the recent version of Life Lexicon redefined "track", and now it mentions something called "active signal object". There's no such thing as "active signal object". I believe this change constitutes damage.

I believe the old definition of "track" (quoted above) is helpful and accurate, and it should be reverted back to the form it was back in 2002, without any modifications.
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » July 18th, 2023, 10:51 pm

Currently 'elementary conduit' is defined as follows:
Life Lexicon (2023-06-24) wrote::elementary conduit A conduit with no recognizable active signal stage besides its input and output. An early example still very commonly used is Buckingham's BFx59H, which transforms a B-heptomino into an inverted Herschel in 59 ticks. The BFx59H elementary conduit is a component in many of the original universal toolkit of Herschel conduits. An extension of the same naming convention is used for elementary conduits, with the first and last letters of the name specifying the input and output signal objects. As with Herschels, an arbitrary orientation and center point is chosen for each object. "Fx" means the signal moves forward and produces a mirror-image output. See Herschel conduit for further details.

Theoretically an elementary conduit may become a composite conduit, if another conduit can be found that shares the beginning or end of the conduit in question. In practice this happens only rarely, because many of the most likely branch points have already been identified: glider (G), LWSS (L) or MWSS (M), Herschel (H), B-heptomino (B), R-pentomino (R), pi-heptomino (P), queen bee shuttle (Q), century or bookend (C), dove (D), and wing (W). A Herschel descendant might qualify, due to the elementary conduit that can be seen in the p184 gun. However, there are very few simple conduits that produce Herschel descendants without Herschels, so in practice this is not a useful branch point.
My comments:
  • In the first sentence, 'active signal' / 'active signal stage' is not a well-understood common term, as far as I can tell. (I assume that the intended meaning can be conveyed here by writing 'active object stage' instead of 'active signal stage', because the definition of 'conduit' uses the term 'active object' in a consistent way.)
  • In the part '...the input and output signal objects.', 'signal object' is not a well-understood common term, as far as I can tell. (I assume that the intended meaning can be conveyed here by writing either '...the input and output active objects.' or '...the input and output objects.')
  • The sentence '"Fx" means the signal moves forward and produces a mirror-image output.' is incorrect, because this sentence confuses the signal itself with the active object carrying the signal. (In this context, a signal is a moving information carried by some active object. Unlike the active object, the signal does not have a geometric shape. Therefore 'mirror image of a signal' is not meaningful.) I assume the intended meaning can be conveyed by writing '"Fx" means the active object moves forward and produces a mirror-image output.'
Proposed replacement (my changes are highlighted):
:elementary conduit A conduit with no recognizable active object stage besides its input and output. An early example still very commonly used is Buckingham's BFx59H, which transforms a B-heptomino into an inverted Herschel in 59 ticks. The BFx59H elementary conduit is a component in many of the original universal toolkit of Herschel conduits. An extension of the same naming convention is used for elementary conduits, with the first and last letters of the name specifying the input and output objects. As with Herschels, an arbitrary orientation and center point is chosen for each object. "Fx" means the active object moves forward and produces a mirror-image output. See Herschel conduit for further details.

Tracks made out of elementary conduits can be used to transfer signals through the Life universe, making elementary conduits useful in engineered signal-processing patterns.

Theoretically an elementary conduit may become a composite conduit, if another conduit can be found that shares the beginning or end of the conduit in question. In practice this happens only rarely, because many of the most likely branch points have already been identified: glider (G), LWSS (L) or MWSS (M), Herschel (H), B-heptomino (B), R-pentomino (R), pi-heptomino (P), queen bee shuttle (Q), century or bookend (C), dove (D), and wing (W). A Herschel descendant might qualify, due to the elementary conduit that can be seen in the p184 gun. However, there are very few simple conduits that produce Herschel descendants without Herschels, so in practice this is not a useful branch point.
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » July 19th, 2023, 4:11 pm

confocaloid wrote:
July 18th, 2023, 10:51 pm
[proposed removal of most uses of the term "signal" from "track" and "elementary conduit" Life Lexicon entries]
I'm not sure if you might continue with further proposals along these lines, for other Life Lexicon entries. Based on a text search of the multi-page HTML version of the Lexicon, the term "signal" is used 274 times in the current version. I think the majority of those uses won't be objectionable to anybody.

However, about a hundred of the mentions of "signal" appear to me to be similar to the meaning used in the two definitions above. Below are some excerpts from various definitions. I don't claim that the excerpts are complete or representative, just that they all contain the word "signal". I skipped over quite a few other uses that seemed very unlikely to be controversial.

To be clear, I'm not suggesting that anyone should work on editing all of these definitions. My preference would be to maybe add a sentence or two to the definition of "signal", to clarify that a commonly understood meaning of "signal" is "active object moving through circuitry" (or something along those lines). With that addition, I personally think all of these other definitions are fine as they are.
:asynchronous Indicates that precise relative timing is not needed for two or more input signals entering a circuit, or two or more sets of gliders participating in a glider synthesis. In some cases the signals or sets of gliders can arrive in any order at all - i.e., they have non-overlapping effects.

However, in some cases such as slow salvo constructions, there is a required order for some of the incoming signals. These signals can still be referred to as "asynchronous" because the number of ticks between them is infinitely adjustable: arbitrarily long delays can be added with no change to the final result. Compare synchronized.

:bait An object in a converter, usually a small still life, that is temporarily destroyed by an incoming signal, but in such a way that a usable output signal is produced. In general such a converter produces multiple output signals (or a signal splitter is added) and one branch of the output is routed to a factory mechanism that rebuilds the bait object so that the converter can be re-used.

:beehive stopper A Spartan logic circuit discovered by Tanner Jacobi on 12 May 2015. It converts an input glider signal into a beehive, in such a way that the beehive can cleanly absorb a single glider from a perpendicular glider stream. The circuit can't be re-used until the beehive "bit" is cleared by the passage of at least one perpendicular input.

:block factory Any factory circuit that produces a block in response to an input signal. For a useful high-clearance example see keeper.

:B track A track for B-heptominoes. A B-heptomino becomes a Herschel plus a block in twenty generations, so this term was nearly synonymous with Herschel track until the discovery of elementary conduits that convert a B directly to another B, or to some other non-Herschel signal output. See for example BRx46B.

:CatForce An optimized search program written by Michael Simkin in 2015, using brute-force enumeration of small Spartan objects in a limited area, instead of a depth-first tree search. One major purpose of CatForce is to find glider-constructible completions for signal conduits. An early CatForce discovery was the B60 conduit, which enabled a record-breaking new glider gun.

:century-to-glider converter Any signal circuit that accepts a century as input and produces a clean output glider. For example, in November 2017 Adam P. Goucher noticed that this previously known C-to-G converter can replace the century eater in Paul Callahan's bistable switch, producing an extra glider output.

:channel A lane or signal path used in construction circuitry. Until the invention of single-channel construction arms, signals in a channel would usually be synchronized with one or more coordinated signals on other paths, as in the Gemini, which used twelve channels to run three construction arms simultaneously, or the 10hd Demonoid which needed only two channels. See also Geminoid.

:circuit Any combination of conduits or converters that moves or processes an active signal. This includes components with multiple states such as period multipliers or switches, which can be used to build guns, logic gates, universal constructors, and other computation or construction circuitry.

:clearance In signal circuitry, the distance from an edge shooter output lane to the last unobstructed lane adjacent to the edge-shooter circuitry. For example, an Fx119 inserter has an unusually high 27hd clearance.

Also, oscillator and eater variants may be said to have better clearance if they allow gliders or other signals to pass closer to them than the standard variant allows. The following high-clearance eater1 variant by Karel Suhajda allows gliders to pass one lane closer on the southeast side, than is allowed by the standard fishhook shape.

:composite conduit A signal-processing conduit that can be subdivided into two or more elementary conduits.

:CP semi-Snark A period-multiplying colour-preserving signal conduit found by Tanner Jacobi in October 2017, producing one output glider for every two input gliders. It is made by replacing one of the eaters in a Snark with a catalyst found using Bellman. The catalyst causes the formation of a tub which requires a second glider to delete. However, this adds 5 ticks to the repeat time, so that it becomes 48. This is still 3 ticks faster than the CC semi-Snark.

:dirty Opposite of clean. A reaction which produces a large amount of complicated junk which is difficult to control or use is said to be dirty. Many basic puffer engines are dirty and need to be tamed by accompanying spaceships in order to produce clean output. Similarly, a dirty conduit is one that does not recover perfectly after the passage of a signal; one or more extra ash objects are left behind (or more rarely a catalyst is damaged) and additional signals must be used to clean up the circuit before it can be re-used.

:dormant An object that is either stable or oscillates without producing any output, until it is triggered by an appropriate signal, which then produces some desired action. For example, freeze-dried objects are dormant until the arrival of a particular glider.

:dr Short identifier for Dean Hickerson's 'drifter' search program, used at various times to find wires, eaters, higher-period billiard table configurations, and related signal-carrying and signal-processing mechanisms. See also drifter.

:eater1 ... This eater can be constructed using a simple two-glider collision, as shown in stamp collection. It is often modified in various ways, or welded to other objects, to allow tighter packing of circuits or to allow a signal stream to pass close by. See clearance for an eater1 variant that is 1hd shorter to the southeast than the standard fishhook form. An eater1 can also be used as a 90-degree one-time turner.

:edge shooter A gun or signal circuit that fires its gliders (or whatever) right at the edge of the pattern, so that it can be used to fire them closely parallel to others. This is useful for constructing complex guns. Compare glider pusher, which can in fact be used for making edge shooters.

:elbow Depending on context, this term may refer to a signal elbow or a construction elbow. See also elbow ladder.

:elementary conduit (quoted above)

:F171 An elementary conduit, the seventeenth Herschel conduit, discovered by Brice Due in August 2006 in a search using only eaters as catalysts...
The central eater in the group of three to the northwest can be removed to release an additional glider output signal on a transparent lane.

:first natural glider The glider produced at T=21 during the evolution of a Herschel. This is the most common signal output from a Herschel conduit.

:fuse A wick burning at one end... A fuse can burn arbitrarily slowly, as demonstrated by the example Blockic fuse below. A signal, alternating between glider and MWSS form, travels up and down between two rows of blocks in a series of one-time turner reactions.

:glider stopper A Spartan logic circuit discovered by Paul Callahan in 1996. It allows a glider signal to pass through the circuit, leaving behind a beehive that can cleanly absorb a single glider from a perpendicular glider stream.

:G-to-H A converter that takes a glider as an input signal and produces a Herschel output, which can then be used by other conduits. G-to-Hs are frequently used in stable logic circuitry. Early examples include Callahan G-to-H, Silver G-to-H, and p8 G-to-H for periodic circuits. A more compact recent example is the syringe.

:Herschel (stabilizes at time 128) The following pattern which occurs at generation 20 of the B-heptomino... Herschels are one of the most versatile types of signal in stable circuitry. R-pentominoes and B-heptominoes naturally evolve into Herschels, and converters have also been found that change pi-heptominoes and several other signal types into Herschels, and vice versa. See elementary conduit.

:Herschel circuit A series of Herschel conduits or other components, connected by placing them so that the output Herschels from early conduits become the input Herschels for later conduits. Often the initial component is a converter accepting some other signal type as input - usually a glider, in which case a syringe is most commonly used.

:Herschel loop A cyclic Herschel track. Although no loop of length less than 120 generations has been constructed it is possible to make oscillators of smaller periods by putting more than one Herschel in a higher-period track. In this way oscillators, and in most cases guns, of all periods from 54 onwards can now be constructed (although the p55 case is a bit strange, shooting itself with gliders in order to stabilize itself). A mechanism for a period-52 loop was found in April 2018, but it includes a stage where the signal is carried by a triplet of gliders so it may not be considered to be a pure Herschel loop. The missing period, 53, is a difficult case simply because 53 is prime and so no small sparkers or reflectors are available.

:Herschel stopper A method of cleanly suppressing a Herschel signal with an asynchronous boat-bit, discovered by Dean Hickerson. Here a ghost Herschel marks the location of the output signal, in cases where the boat-bit is not present. Other boat-bit locations that allow for clean suppression of a Herschel are also known.

:Herschel-to-glider The largest category of elementary conduit. Gliders are very common and self-supporting, so it's much easier to find these than any other type of output signal. A large collection of these H-to-G converters has been compiled, with many different output lanes and timings. These can be used to synchronize multiple signals to produce gun patterns or complex logic circuitry.

:highway robber Any mechanism that can retrieve a signal from a spaceship lane while allowing spaceships on nearby lanes to pass by unaffected. In practice the spaceship is generally a glider. The signal is removed from the lane, an output signal is generated elsewhere, and the highway robber returns to its original state.

:intermittent stream ... Intermittent streams can be used to transmit signals, where holes in the stream can also convey information. For example, the stream can be processed by an inverter having the same period

:Karel's p15 (p15) An oscillator discovered by Karel Suhajda on December 11, 2002. It consists of a period 15 rotor supported by the domino spark of a pentadecathlon. It provides accessible sparks that can be used to perturb reactions or thin signal streams.

:line-mending reaction (*this definition is out of date for other reasons*)
No simple mechanism is known to mend the gap which lies completely on one side of the line. However, it is technically possible to use construction arm technology to push objects through the gap to build and trigger a seed for the required synchronized signals on the other side.

:NW31 One of the most common stable edge shooters... The edge shooter output at the top has no additional clearance, so its use in creating convoys is limited: it can only add gliders on the outermost lanes of an existing glider salvo. Like the beehive version of SW-2, either output can be used to build logical OR gates, where multiple input signal paths are merged onto the same output path.

:overclocking A term used when a circuit can accept a signal at a specific period which it cannot accept at a higher period. A syringe is a simple example.

:over-unity reaction An important concept in gun and macro-spaceship construction. To be a good candidate for building one of these types of patterns with a new period or speed, a stationary reaction (for a gun) or a moving reaction (for a macro-spaceship) must be able to produce some number of output signals, strictly greater than the number of input signals required to maintain the reaction. The extra signal becomes a gun's output stream, or may be used in a variety of ways to construct the supporting track for a macro-spaceship. By implication, "over-unity" refers to the ratio of output signals to input signals.

If all signal outputs must be used up to sustain a stationary reaction, a high-period oscillator may still be possible. See emu for example.

:p8 G-to-H A small periodic variant of a stable two-glider-to-Herschel component found by Paul Callahan in November 1998 and used in the Callahan G-to-H, Silver reflector and Silver G-to-H. The minimum repeat time is 192 ticks, though some lower periods such as 96 are possible via overclocking. Here a ghost Herschel marks the output signal location...

:periodic For circuit mechanisms, "periodic" is the opposite of p1 or stable. Periodic circuits necessarily contain oscillators, and therefore they can generally only accept input signals that are synchronized to the combined period of those oscillators (but see universal regulator).

For signal streams, "periodic" means that signals will only be present in the stream at one out of every n ticks, where n is the period of the stream. In a periodic intermittent stream there may be gaps, so that signals do not always appear at every nth tick. However, if a signal does appear, its distance measured in ticks from previous and future signals will always be an exact multiple of n.

:period multiplier A term commonly used for a pulse divider, because dividing the number of signals in a regular stream by N necessarily multiplies the period by N. The term "period multiplier" can be somewhat misleading in this context, because most such circuits can accept input streams that are not strictly periodic.

:permanent switch A signal-carrying circuit that can be modified so that it cleanly absorbs any future signals instead of allowing them to pass. Optionally there may be a separate mechanism to restore the circuit to its original function.

:pseudo Opposite of true. A gun emitting a period n stream of spaceships (or rakes) is said to be a pseudo period n gun if its mechanism oscillates with a period greater than n. This period will necessarily be a multiple of n. If the base mechanism's period is instead a fraction of n, then a period multiplier must also be present which is considered to be part of the mechanism, and the gun as a whole is still a true period gun. For example, a filter may be used on a lower-period gun to produce a compound gun such as the true p48 gun.

Pseudo period n glider guns are known to exist for all periods greater than or equal to 14, with smaller periods being impossible. All known p14 guns are pseudo guns requiring several signal injections, so they are quite large. The following smaller example is a pseudo period 123 gun, interleaving the streams from two true period 246 guns ...

:quadri-Snark A period-multiplying colour-preserving signal conduit found by Tanner Jacobi in October 2017, producing one output glider for every four input gliders.

:racetrack A pattern in which a signal makes its way in a loop through an "obstacle course" of reactions in order to demonstrate various ways that the signal can be reflected, temporarily stored, and converted.

:reanimation A reaction performed by a convoy of spaceships (or other moving objects) which converts a common stationary object into a glider without harming the convoy. This provides one way for signals that have been frozen in place by some previous reaction to be released for use.

:recovery time The number of ticks that must elapse after a signal is sent through a conduit, before another signal can be safely sent on the same path. In general, a lower recovery time means a more useful conduit. For example, the Snark's very low recovery time allowed for the creation of oscillators with previously unknown periods, 43 and 53.

:repeat time The minimum number of generations that is possible between the arrival of one object and the arrival of the next. This term is used for things such as reflectors or conduits where the signal objects (gliders or Herschels, for example) will interact fatally with each other if they are too close together, or one will interact fatally with a disturbance caused by the other. For example, the repeat time of Dave Buckingham's 59-step B-heptomino to Herschel conduit (shown under conduit) is 58.

:self-supporting A type of pattern, specifically a macro-spaceship, that constructs signals or tracks or other scaffolding to assist its movement, but does not contain complete information about its own structure. Examples include the Caterpillar, Centipede, half-baked knightship, waterbear, and the Caterloopillars. Caterpillar has been used as a general term for self-supporting spaceships, but it is not very appropriate for the HBKs.

:semi-Snark Any small stable signal conduit that produces one output glider for every two input gliders, with a 90 degree reflection. These can act as period-doublers for any glider stream whose period is at least equal to their repeat time, and so adding one of these to a single glider gun often results in a pattern much smaller than the older technology of crossing the output of two guns.

:signal elbow A conduit with signal output 90 degrees from its input. This term is commonly used only for signal wires, particularly 2c/3 signals. A Snark could reasonably be called a "glider elbow", but glider reflector is the standard term. A signal elbow with a recovery time less than 20 ticks would enable a trivial proof that Conway's Life is omniperiodic.

:speed booster Any mechanism which allows a signal (indicated by the presence or absence of a spaceship) to move faster than the spaceship could travel through empty space.

:splitter A signal converter that accepts a single input signal and produces two or more output signals, usually of the same type as the input. An older term for this is fanout, or "fanout device".

:stable pseudo-Heisenburp A multi-stage converter constructed by Dave Greene in January 2007, using a complex recipe found by Noam Elkies to insert a signal into a 2c/3 wire. The wire's high transmission speed allows a signal from a highway robber to catch up to a salvo of gliders.

:staged recovery A type of signal-processing circuit where the initial reaction between catalysts an incoming signal results in an imperfect recovery. A catalyst is damaged, destroyed completely as in a bait reaction, or one or more objects are left behind that must be cleaned up before the circuit can be reused. In any of these three cases, output signals from the circuit must be used to complete the cleanup.

:stop and restart A type of signal circuit where an input signal is converted into a stationary object, which is then re-activated by a secondary input signal. This can be used either as a memory device storing one bit of information, or as a simple delay mechanism. In the following January 2016 example by Martin Grant, a ghost Herschel marks the output signal location, and a "ghost beehive" marks the location of the intermediate still life.

:switch A signal-carrying circuit that can send output signals to two or more different locations, depending on the state of the mechanism. These may be toggle circuits, where the state of the switch changes after each use, or permanent switches that retain the same state through many uses until a change is made with a separate signal.

:synchronized Indicates that precise relative timing is required for two or more input signals entering a circuit, or two or more sets of gliders participating in a glider synthesis. Compare asynchronous. See also salvo and slow glider construction.

:syringe A small stable converter found by Tanner Jacobi in March 2015, accepting a glider as input and producing an output Herschel As of June 2018 it is the smallest known converter of this type, so it is very often used to handle input gliders in complex signal circuitry, as described in Herschel circuit.

:track (quoted above)

:transparent In signal circuitry, a term used for a catalyst that is completely destroyed by the passing signal, then rebuilt. Often (though not always) the active reaction passes directly through the area occupied by the transparent catalyst, then rebuilds the catalyst behind itself, as in the transparent block reaction. See also transparent lane.

:tremi-Snark A colour-preserving period-multiplying signal conduit found by Tanner Jacobi on 7 September 2017, producing one output glider for every three input gliders.

:trigger A signal, usually a single glider, that collides with a seed constellation to produce a relatively rare still life or oscillator, or an output spaceship or other signal.

:one-time turner ... A one-time turner reaction can be used as part of a glider injection mechanism, or as a switching mechanism for a signal.

:twin bees shuttle (p46) Found by Bill Gosper in 1971, this was the basis of all known true p46 guns, and all known p46 oscillators except for glider signal loops using Snarks, until the discovery of Tanner's p46 in 2017.

:universal regulator A universal regulator can allow two complex circuits to interact safely, even if they have different base periods. For example, signals from a stable logic circuit could be processed by a period-30 mechanism, though the precise timing of those signals would change in most cases.

:wire A repeating stable structure, usually fairly dense, that a signal can travel along without making any permanent change. Known wires include the diagonal 2c/3 wire, and orthogonal lightspeed wire made from a chain of beehives. Diagonal lightspeed wires are known, but the required signals are fairly complex and have no known glider synthesis. (*this definition is out of date for other reasons*)

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » July 19th, 2023, 10:45 pm

dvgrn wrote:
July 19th, 2023, 4:11 pm
My preference would be to maybe add a sentence or two to the definition of "signal", to clarify that a commonly understood meaning of "signal" is "active object moving through circuitry" (or something along those lines). With that addition, I personally think all of these other definitions are fine as they are.
In my opinion, in a large fraction of those uses, the word 'signal' does not contribute anything to the clarity of the statement. Jargon can be distracting; the best jargon is no jargon. When it is possible to avoid use of 'signal', my preference would be to avoid use of 'signal'. Frequently, the word is ambiguous, and its addition/removal does not change the intended meaning.

Especially in definitions of other terms, I would try to avoid use of the word 'signal' when possible, for the sake of clarity.
I would prefer to transform Signal into a proper disambiguation page, listing all existing meanings separately from each other.
dvgrn wrote:
July 19th, 2023, 4:11 pm
Based on a text search of the multi-page HTML version of the Lexicon, the term "signal" is used 274 times in the current version. I think the majority of those uses won't be objectionable to anybody.

However, about a hundred of the mentions of "signal" appear to me to be similar to the meaning used in the two definitions above. Below are some excerpts from various definitions. I don't claim that the excerpts are complete or representative, just that they all contain the word "signal". I skipped over quite a few other uses that seemed very unlikely to be controversial.
I think the quoted definition of 'racetrack' agrees with my understanding of this meaning of 'signal'. In each of patterns 'p30-racetrack.rle', 'p46racetrack.rle', there are many gliders -- however, in each of those two patterns, there is exactly one object I would want to describe as a signal-carrying object (or, informally, to describe as a "signal").

Below are my complaints about a number of quoted entries, from what I noticed on a single pass.
  • In the entry 'beehive stopper', the word 'signal' is redundant. 'It converts an input glider into a beehive, ...' is clearer.
  • In the entry 'B track', the word 'signal' is redundant. '... to another B, or to some other non-Herschel output.' is clearer.
  • In the entry 'CatForce', the word 'signal' is redundant. '... glider-constructible completions for conduits.' is clearer.
  • In the entry 'century-to-glider converter', the word 'signal' is redundant. 'Any circuit that accepts...' is clearer.
  • In the entry 'circuit', '...that moves or processes an active object.' would be easier to understand.
  • In the entry 'composite conduit', 'signal-processing' is a redundant distraction. 'A conduit that can be subdivided into two or more elementary conduits.' is easier to comprehend.
  • In the entry 'CP semi-Snark', 'signal' is redundant.
  • In the entry 'edge shooter', 'signal' is redundant.
  • For 'elementary conduit', see my previous post.
  • In the entry 'F171', 'signal' is a redundant distraction. '... to release an additional output glider on a transparent lane.' would be clearer.
  • In the entry 'G-to-H', I think it would be clearer to write 'A converter that takes a glider input and produces a Herschel output, which can then be used by other conduits. G-to-Hs are frequently used in stable circuitry. ...'
  • In the entry 'Herschel stopper', I think both occurrences of 'signal' are redundant.
  • In the entry 'p8 G-to-H', I think 'Here a ghost Herschel marks the output location...' would be clearer.
  • In the entry 'quadri-Snark', 'signal' is redundant.
  • In the entry 'semi-Snark', 'signal' is redundant.
  • In the entry 'splitter', I think it would be clearer to write 'A converter that accepts a single input object and produces two or more output objects, usually of the same type as the input.'
  • In the entry 'staged recovery', 'signal-processing' is redundant.
  • In the entry 'switch', 'signal-carrying' is redundant.
  • In the entry 'transparent', 'completely destroyed by the passing signal' should be replaced by 'completely destroyed by the passing reaction'.
  • In the entry 'tremi-Snark', 'signal' is redundant.
  • In the entry 'twin bees shuttle', 'signal' is redundant. It would be clearer to write '... except for glider loops using Snarks, ...'
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » July 21st, 2023, 8:30 am

confocaloid wrote:
July 19th, 2023, 10:45 pm
Below are my complaints about a number of quoted entries, from what I noticed on a single pass.
Here's the minimum that will have to happen, for a new version of the Life Lexicon to be rolled out, implementing any or all of these suggested changes.

Someone, or preferably several people, will have to
  • review every existing definition -- not just the ones that contain the word "signal". Any definitions that are now misleading or no longer correct will have to be re-worked until they are true again (and hopefully uncontroversial).
  • retrieve the build scripts from the source distribution, and get them working again to build a new set of the four different packaged Life Lexicon releases available at https://conwaylife.com/ref/lexicon.
  • run the script in the current Golly distribution to build the different versions of the Life Lexicon that are embedded in the releases for various platforms. (There's a separate web version hiding in /gui-common/Help/Lexicon/.)
  • test all of these variants of the new Lexicon release.
  • roll out the new version(s) of Golly, and move the Lexicon files to conwaylife.com/ref/lexicon.
This does not include writing definitions for notable new discoveries since 2018. There are quite a few of these.

It doesn't seem to make sense to attempt a few small targeted changes to the Lexicon. The new version will have to get a new date stamp to differentiate it from previous versions. "2023" or "2024" or whatever would imply that the information in the new release is correct as of that date. Until a complete update happens, it seems better to keep the current "2018 July 2" date stamp.

I'm hoping that the new Lexicon release will really be able to bring it back up to date, well enough to last for another half-decade or so. That will mean that someone is going to have to do a lot of very careful review and update work, that really at this point hasn't even started yet.

I've been thinking that I might be able to find time for a new Lexicon release sometime in 2024. I certainly won't be able to do it before then, myself. Of course the Life Lexicon isn't "mine", I'm just the current self-appointed gatekeeper -- and I've been looking to hand over the Head Lexicographer hat to someone else for many years now.

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » August 20th, 2023, 6:07 pm

https://conwaylife.com/w/index.php?titl ... did=134074
confocaloid wrote:
July 19th, 2023, 10:45 pm
  • In the entry 'transparent', 'completely destroyed by the passing signal' should be replaced by 'completely destroyed by the passing reaction'.
Current text:
LifeWiki/Transparent object wrote:In signal circuitry, a transparent object is one that is completely destroyed by the passing signal, then rebuilt. Often (though not always) it is associated with a transparent debris effect, when an active reaction destroys a still life, then later, having passed through the place occupied by still life, recreates the still life in its original position.
Proposed rewrite:
In reusable circuitry, a transparent object is one that is completely destroyed by the passing reaction, then rebuilt. Often (though not always) it is associated with a transparent debris effect, when an active reaction destroys a still life, then later, having passed through the place occupied by still life, recreates the still life in its original position.
Also see 2016 version of related entries in https://web.archive.org/web/20160304131 ... /lex_t.htm
In the entries beginning with "transparent", the discussion is about reactions. There were no mentions of "signals".
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » August 20th, 2023, 11:32 pm

It really does seem like four polite requests to stop removing the term "signal" from LifeWiki articles, for the duration of the current open discussion, should have been enough. Common courtesy between two disagreeing LifeWiki editors should have dictated a temporary cease-fire, without me having to repeat my request so many times.

If I go ahead and revert your change, on the grounds that I honestly believe you're doing damage to consistent valid uses of "signal" in the current LifeWiki, that would probably be the beginning of a pointless edit war. We've almost completely avoided those on the LifeWiki to date, so I won't be starting one now.

Therefore, please revert the above change yourself. You can perfectly well put it back once you have demonstrated that there's broad community support for making these kinds of changes. So far there only seems to be evidence that a disagreement exists.

Life Lexicon changes
Given the current demonstrated level of support for your restrictive "confocaloid-signal" interpretation of signal, I'm not currently inclined to check in any definition changes involving the term "signal" to the Life Lexicon in Golly, or to conwaylife.com/ref/lexicon. Obviously if there's a sudden groundswell of support for your changes from the community, I'll happily reconsider -- but for the time being, that doesn't seem to be happening.

As I mentioned above, to get the Life Lexicon ready for a 2023 or 2024 date-stamped release, there's a large amount of work that will have to be done. Changes to existing definitions that contain the word "signal" seem to me, at best, to be at the very bottom of the priority list. At worst, they're a significant distraction from the job of writing a whole lot of needed new definitions that we can all agree on.

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » August 20th, 2023, 11:45 pm

dvgrn wrote:
August 20th, 2023, 11:32 pm
Given the current demonstrated level of support for your restrictive "confocaloid-signal" interpretation of signal, [...]
FWIW it's what is written in the LifeWiki entry signal, and not "my interpretation":
LifeWiki/Signal wrote:A signal is the movement of information through the Life universe. Signals can be carried by spaceships, fuses, drifters, or conduits. Spaceships can only transfer a signal at the speed of the spaceship, while fuses can transfer a signal at speeds up to the speed of light.

In practice, many signals are encoded as the presence or absence of a glider (or other spaceship) at a particular point at a particular time. Such signals can be combined by the collision of gliders to form logic operations such as AND, OR, and NOT gates. Signals can be duplicated using glider duplicators or other fanout devices, and can be used up by causing perturbations on other parts of the Life object.
I'm very unlikely to start referring to that as "my interpretation" -- that's the common meaning of the term.
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » August 21st, 2023, 12:07 am

dvgrn wrote:
August 20th, 2023, 11:32 pm
Therefore, please revert the above change yourself. ...
Done -- although now the entry confuses signals with reactions, which is bad.
It is more meaningful to say that a transparent object is one that is completely destroyed by the passing reaction, then rebuilt. Signals are irrelevant here.

For the record, my attempt to fix the wording was after this related edit -- feel free to review both.
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » August 21st, 2023, 12:52 am

confocaloid wrote:
August 20th, 2023, 11:45 pm
FWIW it's what is written in the LifeWiki entry signal, and not "my interpretation"...
I'm very unlikely to start referring to that as "my interpretation" -- that's the common meaning of the term.
A single common meaning of the term "signal" in the context of Life, in my opinion, does not actually exist. Like many other terms, "signal" is used to mean many different things in different contexts. Some meanings are more useful than others.

It's perfectly possible to interpret the current LifeWiki definition of "signal", exactly as it's written, to mean "dvgrn-signal" rather than "confocaloid-signal". In fact, if I had had any inkling that it was even possible to read that definition as narrowly as you're insisting on reading it, I would have added another quick sentence back in 2018, so that it was clearer that it shouldn't be interpreted only in that way!

In any case, the current definition of "signal" in the LifeWiki is not set in stone, any more than any of the other definitions that you're currently trying to change. It's theoretically possible that a slight adjustment to the definition of "signal" might make it better reflect actual common usage of the term, without increasing the total amount of confusion in the world.

-- Or maybe you're right after all, of course. Hopefully at some point soon we'll find out what a larger fraction of the community thinks.

Elsewhere you've used Conway's Game of Life: Mathematics and Construction as an authority. Have a look at the bottom of page 209, section 7.8.2:
... it is not known how to quickly or easily convert that junk into one of our standard signals like a Herschel or a glider.
Figure 8.33, page 243:
Figure 8.33: Some conduits that transform various types of signals (gliders, Herschels, and B-heptominoes) into each other more quickly than standard conduits.
I don't believe I wrote either of those sentences, though of course things that I did write may have had an influence. Both of these quotes use the "dvgrn-signal" usage of "signal", not the "confocaloid-signal" definition that you have claimed is the only valid definition that should be mentioned in the signal article.

The point is that the "dvgrn-signal" usage is very widespread. It would be an enormous amount of work to dig it out of everywhere that it currently exists. A lot of existing uses like the ones in the Life textbook really can't be changed at this point.

It wouldn't be so bad doing a lot of work to fix actual mistakes. But what if it's really this limited "confocaloid-signal" definition that you've latched on to, that is the real mistake in practice? What if insisting on trying to make all these changes will actually cause more confusion in the end, because you won't actually be able to eradicate the "dvgrn-signal" common usage?

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » August 21st, 2023, 1:34 am

dvgrn wrote:
August 21st, 2023, 12:52 am
Elsewhere you've used Conway's Game of Life: Mathematics and Construction as an authority.
Note that in that forum post, I referred specifically to Chapter 3, that discusses oscillators, introduces reflectors, and so on. Oscillators and glider reflectors are simpler than signals.

You're referring to later parts of the book, which rely on more complicated concepts and use more complicated jargon.
dvgrn wrote:
August 21st, 2023, 12:52 am
Have a look at the bottom of page 209, section 7.8.2:
... it is not known how to quickly or easily convert that junk into one of our standard signals like a Herschel or a glider.
Would it make difference for the meaning of "transparent object", if a so-called "standard signal" like Herschel was replaced by a "non-standard signal" like the active region hassled in Karel's p177?
Would it make difference for the meaning of "transparent object", if the passing reaction was not related to any signal at all, and if it was not reusable in any known way outside one specific oscillator?
dvgrn wrote:
August 21st, 2023, 12:52 am
The point is that the "dvgrn-signal" usage is very widespread. It would be an enormous amount of work to dig it out of everywhere that it currently exists. A lot of existing uses like the ones in the Life textbook really can't be changed at this point.
I'm not suggesting to "dig it out of everywhere that it currently exists". It is a fact of Life that people will always use things in undocumented ways, both intentionally and unintentionally. That includes using words and phrases in undocumented ways that are not explained either in the Life Lexicon or on LifeWiki.
However, written definitions should be meaningful. If you're defining a concept, then there should be many counterexamples (things that are not examples of the concept).
When it becomes too easy to describe (almost) everything as a "signal", then the word "signal" becomes useless / overused. The same goes for any other overused term/phrase.

The current definition of signal is reasonably meaningful. It says that signals can be "carried by", that they can be "encoded", etc.
I believe the definition should remain as it is -- relaxing it will do more harm than good.
dvgrn wrote:
August 21st, 2023, 12:52 am
It wouldn't be so bad doing a lot of work to fix actual mistakes. But what if it's really this limited "confocaloid-signal" definition that you've latched on to, that is the real mistake in practice? What if insisting on trying to make all these changes will actually cause more confusion in the end, because you won't actually be able to eradicate the "dvgrn-signal" common usage?
Obviously I can't know the future. However, there is a difference between
(a) an attempt to do a specific change, trying to improve something
(b) reverting such an attempt via the rollback functionality, and participating as one of active sides in a long heated argument without clear solution/way out, instead of trying to fix things incrementally.

(The version restored by the rollback added many uses of "signal" in what I consider to be an indiscriminate use of the word. When the word "signal" is too common, it becomes noise.

Also, the version restored by the rollback happened to add misinformation, the way I read the linked edit summary.)
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » August 21st, 2023, 10:01 am

confocaloid wrote:
August 21st, 2023, 1:34 am
The current definition of signal is reasonably meaningful. It says that signals can be "carried by", that they can be "encoded", etc.
The key word for me is "can". It doesn't say "must". If it had ever said anything like that, I would have adjusted the definition long ago (half a decade before this discussion started) -- because that would have made the definition clearly not true, in my view.
confocaloid wrote:
August 21st, 2023, 1:34 am
... there is a difference between
(a) an attempt to do a specific change, trying to improve something
(b) reverting such an attempt via the rollback functionality, and participating as one of active sides in a long heated argument without clear solution/way out, instead of trying to fix things incrementally.
There is certainly a difference between those two things. Are you mentioning it as some kind of indirect criticism of the rollback? I don't do rollbacks lightly, for sure -- this is the first one I've done for a long time. I've been participating in the discussion as a regular LifeWiki editor, not using or mentioning any irrelevant moderator powers -- and a rollback is one of the things that any LifeWiki editor is allowed to do, when that seems like the best option.

The article has long since moved on from its reverted state, and in fact you were able to remove all instances of the term "signal" from it. It's not clear to me why this case was a relevant one to bring up.

For the larger "signal" question, this doesn't seem to be the kind of problem that can be solved by "fixing things incrementally". The opinions involved are diametrically opposed to each other. It seemed to me that community input was needed first, and then a solution could be implemented if and when one showed up.

To try to make incremental progress, would you prefer that I make an experimental adjustment to the "signal" page, similar to what you did here -- and then see what people think of it? I'd be happy to add something like this final sentence to the Life Lexicon and LifeWiki definitions:
A signal is the movement of information through the Life universe. Signals can be carried by spaceships, fuses, drifters, or conduits. Spaceships can only transfer a signal at the speed of the spaceship, while fuses can transfer a signal at speeds up to the speed of light. "Signal" may also refer to any stage of an active object traveling through circuitry.
If people think that the medium-vs.-message distinction is still confusing after that change, I could add another sentence along these lines:
A signal will often but not always correspond to a logical bit of information; the active object itself can be considered to be a minimal type of moving information.

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » August 21st, 2023, 5:36 pm

At least on Wikipedia, rollback is very different from simple "undo" or manual revert. Use of rollback for reverting good-faith changes which you happen to disagree with is likely to be considered misuse of the tool. https://en.wikipedia.org/wiki/Wikipedia:ROLLBACKUSE
dvgrn wrote:
August 21st, 2023, 10:01 am
confocaloid wrote:
August 21st, 2023, 1:34 am
The current definition of signal is reasonably meaningful. It says that signals can be "carried by", that they can be "encoded", etc.
The key word for me is "can". It doesn't say "must". If it had ever said anything like that, I would have adjusted the definition long ago (half a decade before this discussion started) -- because that would have made the definition clearly not true, in my view.
Currently, the LifeWiki entry signal attempts to define a single unambiguous concept. A signal is already defined to be logically distinct from the moving object that carries the signal.
Given that, by definition, signals can be carried by e.g. spaceships, it follows that signals are not those spaceships.
Given that many signals are encoded as the presence or absence of a glider, it follows that signals are not those gliders.

I would prefer to keep the current definition of "signal" unchanged as it is. Any explanation of the existing usage of the word should be clearly separated from the definition of the term.
dvgrn wrote:
August 21st, 2023, 10:01 am
I'd be happy to add something like this final sentence to the Life Lexicon and LifeWiki definitions:
A signal is the movement of information through the Life universe. Signals can be carried by spaceships, fuses, drifters, or conduits. Spaceships can only transfer a signal at the speed of the spaceship, while fuses can transfer a signal at speeds up to the speed of light. "Signal" may also refer to any stage of an active object traveling through circuitry.
I think adding this sentence into the definition of signal would do significant damage. It would be an attempt to combine two different definitions of different meanings of the word 'signal' into one "definition".

If the same sentence is added as a separate paragraph, then that addition will have the effect of converting the LifeWiki page signal into a disambiguation page that lists two or more different meanings of the word "signal". The page will then have to be recognized as a disambiguation page, rather than a single-concept definition.
dvgrn wrote:
August 21st, 2023, 10:01 am
The article has long since moved on from its reverted state, and in fact you were able to remove all instances of the term "signal" from it. It's not clear to me why this case was a relevant one to bring up.

For the larger "signal" question, this doesn't seem to be the kind of problem that can be solved by "fixing things incrementally". The opinions involved are diametrically opposed to each other. It seemed to me that community input was needed first, and then a solution could be implemented if and when one showed up.
I consider that a relevant case to bring up, because you did in fact object to some specific changes to specific LifeWiki pages, even though those changes do not directly address the larger "signal" question. When I edited the "elementary conduit" entry in an attempt to clarify the wording, you objected to that specific change.

I believe my change to "elementary conduit" is an improvement. Currently, my version is there. However, I cannot know whether it will remain there, or if it will be suddenly rollbacked at some point in future and labeled as "damage" or suchlike.
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » August 21st, 2023, 9:34 pm

confocaloid wrote:
August 21st, 2023, 5:36 pm
At least on Wikipedia, rollback is very different from simple "undo" or manual revert.
Thanks for clarifying this. I remember deciding to undo back to the previous version in that case, because there were too many related changes that all subtly depended on what I thought was an invalid assumption. It seemed better to get a fresh start and try really hard to convince you that that it wasn't a good idea to proceed with any further edits based on that assumption.

I use rollbacks so seldom -- pretty close to never, in fact -- that I didn't remember that clicking "rollback" instead of "undo" would not give me an opportunity to put in an edit summary. However, it certainly wasn't an unexplained rollback. I reached out right away with an apology, and continued the conversation with you on the talk page, as follows:
:::::: With apologies, I've rolled back the most recent edit, which had the edit summary
eliminate mentions of "signals" that are not actual signals and "loops" that are not actual signal loops. Different sets of gliders flowing between dependent reflectors belong to different glider streams. Removing the newly coined term "dependent reflector loop", which is confusing for reasons I explained on the talk page; the article is now primarily about dependent reflectors, not about oscillators ...
I thought that there were too many changes at once in that edit, especially the removal of "dependent reflector loop". You yourself later realized that "dependent reflector loop" was an acceptable term after all, so maybe it made some sense to undo that particular change (even if I accidentally did it as a rollback).

Long story short -- there was absolutely no intention of conveying any judgment about the edit, by my use of "rollback" instead of "undo". That part was a complete accident, and I'm sorry if it caused you any concerns. You didn't mention any such concerns at the time, so it would have been hard for me to guess that there was a problem with my method of reverting.
confocaloid wrote:
August 21st, 2023, 5:36 pm
I believe my change to "elementary conduit" is an improvement. Currently, my version is there. However, I cannot know whether it will remain there, or if it will be suddenly rollbacked at some point in future and labeled as "damage" or suchlike.
This being a wiki, none of us can really know if something will be suddenly removed at some point in the future.

However, if you're actually intending to say that you're worried about me specifically rolling back your edits, please take a good look at the very careful series of communications that you've gotten from me so far:

First, I did one rollback -- accidentally, in that I would have preferred to undo -- and immediately apologized for doing that.

Next, you made four edits that removed the term "signal" from places where I had deliberately added it half a decade before, where I believed that the uses of the term were perfectly valid -- even, that they were useful in helping to make clear that the defined objects fell into the category of "signal circuitry".

Each time you made one of these edits, I pointed out that that was what you were doing, but I didn't either roll back or undo a single one of them. I really don't like using either the "undo" or the "rollback" functions under normal circumstances. The one initial rollback was a special case, as I've explained above.

For the fifth removal of "signal", I asked you to revert the edit yourself, until the community could offer some more clarity about which way to go with this definition. You did that right away, with only a very understandable amount of protest.

Thank you for doing that. I was starting to really worry that you would just continue to ignore my requests and continue the same pattern of editing out "signal" before hearing other people's feedback.

I hope this makes it clear that I'm not likely to start arbitrarily rolling back your work. It's probably also clear by this time that I'm just as determined as you are not to let this particular issue be decided solely according to which of us is the most energetic or stubborn LifeWiki editor. Some additional input from the community is needed here.

User avatar
confocaloid
Posts: 2723
Joined: February 8th, 2022, 3:15 pm

Re: Life Lexicon update collection thread

Post by confocaloid » August 23rd, 2023, 11:39 pm

I believe the word 'signal' is overused. It may indeed refer to any stage of an active object traveling through circuitry, but often there is a more clear wording.
Even if one doesn't consider places such as definitions of terms (where ambiguity / informal jargon is obviously undesirable, because definitions should be clear), there are many examples of this.

Another example can be found in the template documentation page Template:Conduit/Doc (permanent link). That page contains part "... when the input signal has a horizontal line of symmetry ..." I think it would be an improvement to change 'signal' to 'pattern' (to be consistent with the rest of the linked page), or to 'object'.

I think many examples of this kind on LIfeWiki could be rewritten/reworded for better clarity, if the issue was not considered controversial.
dvgrn wrote:
August 21st, 2023, 10:01 am
[...]
"Signal" may also refer to any stage of an active object traveling through circuitry.
[...]
confocaloid wrote:
August 21st, 2023, 5:36 pm
[...]
Currently, the LifeWiki entry signal attempts to define a single unambiguous concept. A signal is already defined to be logically distinct from the moving object that carries the signal.
[...]
If the same sentence is added as a separate paragraph, then that addition will have the effect of converting the LifeWiki page signal into a disambiguation page that lists two or more different meanings of the word "signal". The page will then have to be recognized as a disambiguation page, rather than a single-concept definition.
[...]
dvgrn wrote:
August 21st, 2023, 9:34 pm
[...] It's probably also clear by this time that I'm just as determined as you are not to let this particular issue be decided solely according to which of us is the most energetic or stubborn LifeWiki editor. Some additional input from the community is needed here.
127:1 B3/S234c User:Confocal/R (isotropic rules, incomplete)
Unlikely events happen.
My silence does not imply agreement, nor indifference. If I disagreed with something in the past, then please do not construe my silence as something that could change that.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Life Lexicon update collection thread

Post by dvgrn » August 24th, 2023, 2:23 pm

confocaloid wrote:
August 23rd, 2023, 11:39 pm
I believe the word 'signal' is overused. It may indeed refer to any stage of an active object traveling through circuitry, but often there is a more clear wording.
Even if one doesn't consider places such as definitions of terms (where ambiguity / informal jargon is obviously undesirable, because definitions should be clear), there are many examples of this.

Another example can be found in the template documentation page Template:Conduit/Doc (permanent link). That page contains part "... when the input signal has a horizontal line of symmetry ..." I think it would be an improvement to change 'signal' to 'pattern' (to be consistent with the rest of the linked page), or to 'object'.
Not surprisingly, I'd rather keep "signal" the way it is in that case, because it's a nice specific term where "pattern" or "object" would be unnecessarily vague. The input isn't just any old pattern or object, it's specifically an active object traveling through circuitry -- a dvgrn-signal. In many cases it will be a confocaloid-signal too -- or rather, it will carry a confocaloid-signal, without actually being one -- but that's not guaranteed.

This is precisely why it seems like such an awkward idea to limit "signal" to mean only "confocaloid-signal", i.e., abstract information-theory encoded information.

If "signal" is only allowed to mean "confocaloid-signal", then we can't say things like "... when the input signal has a horizontal line of symmetry ...". But if the definition of "signal" includes dvgrn-signals, then the meaning of that phrase remains perfectly clear, and then there might not be any particular reason to change the wording.
confocaloid wrote:
August 23rd, 2023, 11:39 pm
I think many examples of this kind on LIfeWiki could be rewritten/reworded for better clarity, if the issue was not considered controversial.
It seems as if your idea of "clarity" implicitly involves making it clear to everyone that "confocaloid-signal" is the primary definition of "signal". Conversely, it seems important to me to make it clear that the "dvgrn-signal" meaning is also perfectly valid usage, not any kind of "jargon". It probably shouldn't be surprising that pretty much every instance of "signal" that you want to remove would improve clarity from your point of view, but would do damage to clarity from my point of view.

Now, if we could settle the issue of whether "signal" can be safely used on the LifeWiki to mean "active object moving through signal circuitry", then I think I wouldn't have to be so worried about the removal of "signal" from a few places where maybe it's not strictly needed. I totally agree that there are some definitions where "signal" is a little bit redundant.

That said, though, a little bit of redundancy isn't always confusing or undesirable. Quite often I think it can serve as good reinforcement of a concept, in which case it can be absolutely fine and even preferable to leave it in place -- even if there's a more concise alternate wording that really technically means the same thing. I'd say that that's the case for most of the uses of "signal" that you've highlighted so far -- not that it's surprising that I would think that, since I put most of them there deliberately, half a decade ago or so.

Next Step Forward?
I've heard from a few people now that they don't see the point of posting their opinions on any of the currently open threads on the "signal" topic -- just because it's obvious to them that their opinions would get promptly drowned out by what amount to copy/pasted restatements of either my opinion or yours.

On the other hand, the recent "poll" thread on the definition of "dependent reflector" actually got some attention, to the point where it might well be considered to be a statistically significant expression of community opinion.

So that seems like it might be an idea for making some progress on the "signal" question. What do you think of starting a poll thread to collect community sentiment, using a short summary of your proposal and my proposal?

Here's my initial attempt at a poll question. Now, we've both demonstrated by now that we're not too good at guessing what the other person is thinking or what they'll object to, so please feel free to suggest alternate poll items. Maybe we can agree on wording that seems fair to both of us, not slanted towards one opinion or another. Then someone can do a poll post once the wording is settled.

It seems good to me to keep the number of options to a minimum, since the issue really does seem to boil down to two diametrically opposite ideas about how "signal" should be defined and used in LifeWiki articles.
  • The p43 Snark loop contains no signals. Circulating gliders in a context like this should not be referred to as "signals" in LifeWiki articles, because they do not encode any information. The current LifeWiki definition of "signal" is complete and correct as it stands. LifeWiki and Life Lexicon articles should be edited to remove uses of "signal" that imply that the p43 Snark loop can contain signals.
  • The p43 Snark loop contains eight circulating signals. The LifeWiki "signal" article should be adjusted slightly so that it's clear that this is a mainstream use of the term "signal". For example, this sentence could be added: "Signal" can also refer directly to active objects traveling through circuitry -- gliders, Herschels, B-heptominoes, etc.
  • Other. If neither of the above two options looks right, please respond to this thread with a post stating an option that you would vote for.

Post Reply