Catagolue Oddities

For general discussion about Conway's Game of Life.
User avatar
LaundryPizza03
Posts: 1061
Joined: December 15th, 2017, 12:05 am
Location: Unidentified location "https://en.wikipedia.org/wiki/Texas"

Re: Catagolue Oddities

Post by LaundryPizza03 » October 1st, 2020, 1:06 pm

This haul was uploaded under the rule b2i3aceijnry4cj5cy6cn7cs2acekn34cy7c8, which is a non-canonical version of b2i3-kq4cj5cy6cn7cs2-i34cy7c8. Perhaps INT hauls should be auto-canonized.

Code: Select all

x = 4, y = 3, rule = B3-q4z5y/S234k5j
2b2o$b2o$2o!
LaundryPizza03 at Wikipedia

The latest edition of new-gliders.db.txt and oscillators.db.txt have 31150 spaceships and 1205 oscillators from outer-totalistic rules. You are invited to help!

MathAndCode
Posts: 693
Joined: August 31st, 2020, 5:58 pm

Re: Catagolue Oddities

Post by MathAndCode » October 1st, 2020, 5:29 pm

Linear growth patterns "work" in incorrect rules.

Catalogue uses apparent hashes that can be longer than the actual apgcodes.

Pages for replicators in one rule have comments about replicators in different rules.

This rule's rulestring displays incorrectly. (It should be S,B4, not S0,0,0,0,0,0,B0,0,0,0,0,8.)

Period-11,843 guns are apparently replicators. I would get pathological because the period is over 1,000, but I do not understand how it thinks that a gun is a replicator.

The LifeViewers for objects in Marine (example) don't work because they are fed Marine as a rulestring. The correct rulestring for the orientation that the patterns are shown in would be R2,C2,S,B1,N@f7bbc0, but the forum thread uses R2,C2,S4,6-9,B6-8,N@03ddef (which is a 180° rotation of the former), and I'm used to N@03ddef, so if the rulestring is updated, I'd prefer for it to be updated to that if possible, although it might not be possible to rotate all of the patterns and soups by 180°, in which case I will understand. (Of course, I'll also understand if this isn't updated because Marine is a relatively obscure rule.)

In Marine, separate spaceships are counted as one spaceship.

Many of the Life-like rules have dead links to Fano.

Why does Catagolue catalogue the rule where nothing interesting happens because nothing happens or the rule where nothing interesting happens because every generation is the same as before except inverted, so every pattern that is not on a torus or another grid of finite cell count is infinitely oversized?

Is Catagolue a misspelling of catalogue?
Marine has a non-centered Moore neighborhood and at least one elementary gun with a very high period (11,843).
I'm in the process of reducing the glider cost of universal construction.

User avatar
goldenratio
Posts: 124
Joined: July 26th, 2020, 10:39 pm
Location: Exactly where you'd expect

Re: Catagolue Oddities

Post by goldenratio » October 1st, 2020, 6:11 pm

MathAndCode wrote:
October 1st, 2020, 5:29 pm
A bunch of oddities
I can explain some of these:

Oddity 1: Catagolue doesn't check if linear growth apgcodes are actually valid since (as far as I know) they are based on population changes and not what the pattern actually is: http://catagolue.hatsya.com/object/yl5_ ... owth/b3s23

Oddity 3: Comments for an apgcode are shared for all rules.

Oddity 7: Pseudo-objects whose individual objects' histories overlap are counted as one object, due to RROs and other reasons.

Oddity 9: Because someone decided to search it. Literally no other reason.

"Oddity" 10: It's a pun of "catalogue" and GOL (game of life): cataGOLue.
cool epic gun since I have nothing else for a signature

Code: Select all

x = 5, y = 7, rule = B2n3-cq4e5y6c7c/S23-k
2o$2o3$3bo$3b2o$3bo!

MathAndCode
Posts: 693
Joined: August 31st, 2020, 5:58 pm

Re: Catagolue Oddities

Post by MathAndCode » October 1st, 2020, 6:13 pm

goldenratio wrote:
October 1st, 2020, 6:11 pm
Oddity 3: Comments for an apgcode are shared for all rules.
But the example that I gave is not a real apgcode. Even different patterns in the same rule could both have that apgcode.
goldenratio wrote:
October 1st, 2020, 6:11 pm
Oddity 1: Catagolue doesn't check if linear growth apgcodes are actually valid since (as far as I know) they are based on population changes and not what the pattern actually is: http://catagolue.hatsya.com/object/yl5_ ... owth/b3s23
The population the set of cells corresponding to a BLSE does not become infinite in B3S135.
Marine has a non-centered Moore neighborhood and at least one elementary gun with a very high period (11,843).
I'm in the process of reducing the glider cost of universal construction.

User avatar
Saka
Posts: 3598
Joined: June 19th, 2015, 8:50 pm
Location: somewhere
Contact:

Re: Catagolue Oddities

Post by Saka » October 1st, 2020, 10:39 pm

MathAndCode wrote:
October 1st, 2020, 6:13 pm
But the example that I gave is not a real apgcode. Even different patterns in the same rule could both have that apgcode.

The population the set of cells corresponding to a BLSE does not become infinite in B3S135.
Yes, all replicators in a rule are lumped into the single "zz_REPLICATOR" object. Apgcodes don't even have to be real, you could make something up. Catagolue also doesn't check apgcodes for xp0_whatever, xs0_whatever and xq0_whatever.

Catagolue does not check for the legitimacy of yl apgcodes, similar to xp0 and xq0 above. It automatically just puts the BLSE image on the page of that apgcode for any rule.
Period-11,843 guns are apparently replicators.
This is a problem with apgsearch rather than Catagolue. It has looked at the population graph (and possibly other things as well) for the pattern and has determined it is a replicator, and marked it as "zz_REPLICATOR". Occasionally you'll also get PATHOLOGICALS that are just regular objects as well (this is an apgsearch problem, not Catagolue).
Marine...
It has been advised by Calcyman (APGoucher, creator of Catagolue / Apgsearch) that custom rules searched should have a minimum of rotate4reflect symmetry, this is because to obtain the canonical apgcode of a pattern, all reflections and symmetries are tried for the shortest / alphabetically first apgcode

Code: Select all

o3b2ob2obo3b2o2b2o$bo3b2obob3o3bo2bo$2bo2b3o5b3ob4o$3o3bo2bo2b3o3b3o$
4bo4bobo4bo$2o2b2o2b4obo2bo3bo$2ob4o3bo2bo2bo2bo$b2o3bobob2o$3bobobo5b
obobobo$3bobobob2o3bo2bobo!
(Check gen 3)
Add your computer to the Table of Lifeenthusiast Computers!

MathAndCode
Posts: 693
Joined: August 31st, 2020, 5:58 pm

Re: Catagolue Oddities

Post by MathAndCode » October 1st, 2020, 11:04 pm

Saka wrote:
October 1st, 2020, 10:39 pm
It has been advised by Calcyman (APGoucher, creator of Catagolue / Apgsearch) that custom rules searched should have a minimum of rotate4reflect symmetry, this is because to obtain the canonical apgcode of a pattern, all reflections and symmetries are tried for the shortest / alphabetically first apgcode
Ah; thank you for explaining. But why does apgcode use Marine, which LifeViewer does not recognize, instead of a recognizable rulestring?
Marine has a non-centered Moore neighborhood and at least one elementary gun with a very high period (11,843).
I'm in the process of reducing the glider cost of universal construction.

bubblegum
Posts: 392
Joined: August 25th, 2019, 11:59 pm
Location: click here to do nothing

Re: Catagolue Oddities

Post by bubblegum » October 1st, 2020, 11:14 pm

MathAndCode wrote:
October 1st, 2020, 5:29 pm
Linear growth patterns "work" in incorrect rules.

Catalogue uses apparent hashes that can be longer than the actual apgcodes.

Pages for replicators in one rule have comments about replicators in different rules.

This rule's rulestring displays incorrectly. (It should be S,B4, not S0,0,0,0,0,0,B0,0,0,0,0,8.)

Period-11,843 guns are apparently replicators. I would get pathological because the period is over 1,000, but I do not understand how it thinks that a gun is a replicator.

The LifeViewers for objects in Marine (example) don't work because they are fed Marine as a rulestring. The correct rulestring for the orientation that the patterns are shown in would be R2,C2,S,B1,N@f7bbc0, but the forum thread uses R2,C2,S4,6-9,B6-8,N@03ddef (which is a 180° rotation of the former), and I'm used to N@03ddef, so if the rulestring is updated, I'd prefer for it to be updated to that if possible, although it might not be possible to rotate all of the patterns and soups by 180°, in which case I will understand. (Of course, I'll also understand if this isn't updated because Marine is a relatively obscure rule.)

In Marine, separate spaceships are counted as one spaceship.

Many of the Life-like rules have dead links to Fano.

Why does Catagolue catalogue the rule where nothing interesting happens because nothing happens or the rule where nothing interesting happens because every generation is the same as before except inverted, so every pattern that is not on a torus or another grid of finite cell count is infinitely oversized?

Is Catagolue a misspelling of catalogue?
I'll explain all of these:
1) apgsearch is blind to infinite growth cell positioning (after all there is some extra garbage that they produce that isn't part of them). The code to apply the BLSE/GPSE checks the apgcode, not the validity of it.
2) Same as above. apgsearch hashes for population growth at timestamps. (Don't ask me how it does that.)
3) Catagolue is being lazy.
4) These are called lifelib rulestrings. It just so happens that HROT rulestrings in lifelib format suck.
5) apgsearch tracks population growth, being blind to anything else.
6) That's fixable. In the meantime, don't search Marine or anything else without rotate4reflect symmetry.
7) apgsearch is terrible at spaceship spearing... separation. Don't bother.
8) I'll fix that.
9) Somebody decided to search them. Duh.
10) cataGOLue. Sometimes you really need to wait for the tide to turn to sea a pun or two.
Each day is a hidden opportunity, a frozen waterfall that's waiting to be realised, and one that I'll probably be ignoring
bubblegum wrote:
July 2nd, 2020, 8:33 pm
conwaylife signatures are amazing[citation needed]
part-time stator reducer

bubblegum
Posts: 392
Joined: August 25th, 2019, 11:59 pm
Location: click here to do nothing

Re: Catagolue Oddities

Post by bubblegum » October 1st, 2020, 11:17 pm

MathAndCode wrote:
October 1st, 2020, 11:04 pm
Saka wrote:
October 1st, 2020, 10:39 pm
It has been advised by Calcyman (APGoucher, creator of Catagolue / Apgsearch) that custom rules searched should have a minimum of rotate4reflect symmetry, this is because to obtain the canonical apgcode of a pattern, all reflections and symmetries are tried for the shortest / alphabetically first apgcode
Ah; thank you for explaining. But why does apgcode use Marine, which LifeViewer does not recognize, instead of a recognizable rulestring?
why can't i introduce a quote into an edit

Because you specified Marine. Why do cooks never mine their own salt?
Each day is a hidden opportunity, a frozen waterfall that's waiting to be realised, and one that I'll probably be ignoring
bubblegum wrote:
July 2nd, 2020, 8:33 pm
conwaylife signatures are amazing[citation needed]
part-time stator reducer

User avatar
wwei23
Posts: 1566
Joined: May 22nd, 2017, 6:14 pm
Location: The (Life?) Universe

Re: Catagolue Oddities

Post by wwei23 » October 1st, 2020, 11:24 pm

bubblegum wrote:
October 1st, 2020, 11:14 pm
7) apgsearch is terrible at spaceship spearing... separation. Don't bother.
As far as I can tell, the python version usually does much better than the command line version at this.

User avatar
Saka
Posts: 3598
Joined: June 19th, 2015, 8:50 pm
Location: somewhere
Contact:

Re: Catagolue Oddities

Post by Saka » October 1st, 2020, 11:41 pm

bubblegum wrote:
October 1st, 2020, 11:17 pm

Because you specified Marine
To correct / nitpick this a bit, it seems that lemon41625 has been searching marine, not MatHandCode (https://catagolue.hatsya.com/census/xmarine/C1), and they did so using a custom ruletable (because you need ruletables to search custom neighborhood rules as of now), and Catagolue takes the name of the ruleTable and puts it into LifeViewer (which makes sense).

Code: Select all

o3b2ob2obo3b2o2b2o$bo3b2obob3o3bo2bo$2bo2b3o5b3ob4o$3o3bo2bo2b3o3b3o$
4bo4bobo4bo$2o2b2o2b4obo2bo3bo$2ob4o3bo2bo2bo2bo$b2o3bobob2o$3bobobo5b
obobobo$3bobobob2o3bo2bobo!
(Check gen 3)
Add your computer to the Table of Lifeenthusiast Computers!

MathAndCode
Posts: 693
Joined: August 31st, 2020, 5:58 pm

Re: Catagolue Oddities

Post by MathAndCode » October 21st, 2020, 11:06 pm

The twenty-ninth most common object (Big S) apparently has only appeared eleven times in C1. I've seen Catagolue not mention any soups if an object is too common, but I've never seen it give a falsely low number before.
Marine has a non-centered Moore neighborhood and at least one elementary gun with a very high period (11,843).
I'm in the process of reducing the glider cost of universal construction.

wildmyron
Posts: 1466
Joined: August 9th, 2013, 12:45 am

Re: Catagolue Oddities

Post by wildmyron » October 21st, 2020, 11:16 pm

MathAndCode wrote:
October 21st, 2020, 11:06 pm
The twenty-ninth most common object (Big S) apparently has only appeared eleven times in C1. I've seen Catagolue not mention any soups if an object is too common, but I've never seen it give a falsely low number before.
This is related to apgsearch not generally reporting sample soups for very common objects. I can't remember what the threshold is but the original design choice centred on the assumption that for very common objects it would be easy enough to find soups that generated the object manually and therefore recording soups for them would just needlessly increase the haul size. However, due to a quick in sample soup reporting (I think related to pseudo object separation, or possibly linear growth detection) there were some situations where common objects would have sample soups recorded. I don't know if current versions of apgsearch still have this behaviour.
The latest version of the 5S Project contains over 226,000 spaceships. There is also a GitHub mirror of the collection. Tabulated pages up to period 160 (out of date) are available on the LifeWiki.

User avatar
Ian07
Posts: 623
Joined: September 22nd, 2018, 8:48 am

Re: Catagolue Oddities

Post by Ian07 » October 22nd, 2020, 7:14 am

wildmyron wrote:
October 21st, 2020, 11:16 pm
This is related to apgsearch not generally reporting sample soups for very common objects. I can't remember what the threshold is but the original design choice centred on the assumption that for very common objects it would be easy enough to find soups that generated the object manually and therefore recording soups for them would just needlessly increase the haul size. However, due to a quick in sample soup reporting (I think related to pseudo object separation, or possibly linear growth detection) there were some situations where common objects would have sample soups recorded. I don't know if current versions of apgsearch still have this behaviour.
Unfortunately, for me this is sometimes annoying because I can't find the original hauls in which these objects occurred. For example a few days ago I was looking at the b3s23/ikpx2_stdin census and found yl8_1_3_1463956f33241332e72e4a2c9f4ecc61, which has over 300 occurrences but no sample soups. And due to the nature of ikpx2_stdin, what probably happened was that the object appeared several hundred times in a few hauls, and zero times everywhere else, making it difficult to track down. Based on its apgcode it's definitely some sort of c/2 blinker puffer, but I don't know if it's the standard xWSS-based one or something different.

User avatar
calcyman
Posts: 2320
Joined: June 1st, 2009, 4:32 pm

Re: Catagolue Oddities

Post by calcyman » October 22nd, 2020, 9:15 am

Ian07 wrote:
October 22nd, 2020, 7:14 am
Unfortunately, for me this is sometimes annoying because I can't find the original hauls in which these objects occurred. For example a few days ago I was looking at the b3s23/ikpx2_stdin census and found yl8_1_3_1463956f33241332e72e4a2c9f4ecc61, which has over 300 occurrences but no sample soups. And due to the nature of ikpx2_stdin, what probably happened was that the object appeared several hundred times in a few hauls, and zero times everywhere else, making it difficult to track down. Based on its apgcode it's definitely some sort of c/2 blinker puffer, but I don't know if it's the standard xWSS-based one or something different.
Fortunately none of the hauls had been deleted so I was able to retrieve this one manually:

Code: Select all

x = 21, y = 106, rule = B3/S23
8bo3bo$7b3ob3o$6b2obobob2o$5b2o7b2o$6bo7bo$5b2o7b2o$3b2obobo3bobob2o$
4bobob2ob2obobo$b2obo2bobobobo2bob2o$b2obob2obobob2obob2o$bo2bobo2bobo
2bobo2bo$2o2bo11bo2b2o$8b2ob2o$4b2o9b2o$3bo13bo$3b2o11b2o$3b2o11b2o$b
2o15b2o$bo2bo11bo2bo$obob2o9b2obobo$4bobo7bobo$3b2ob2o5b2ob2o$3bo13bo$
6bobo3bobo$2b2o2bob2ob2obo2b2o$5b2o7b2o$9bobo$5bo3bobo3bo$4bo2bo5bo2bo
$5bo9bo2$3b2ob2o5b2ob2o$3b2ob2o5b2ob2o$3bo3bo5bo3bo$2b2o3b2o3b2o3b2o$
2bo5bo3bo5bo$bobo3bobobobo3bobo$5bo9bo$4bobo7bobo$3b2ob2o5b2ob2o$2bob
4o5b4obo$b2o15b2o$o3bob2o5b2obo3bo$4b2obo5bob2o$2obob3o5b3obob2o$7b2o
3b2o$bo17bo$bo17bo2$2o17b2o$2bo15bo$o2bo5b3o5bo2bo$bo2bo3bo3bo3bo2bo$
4bo2bo5bo2bo$5b3o5b3o$2b2obobob3obobob2o$2b2o3bobobobo3b2o$3b2o2bo5bo
2b2o$3bo4bo3bo4bo$3bo6bo6bo$9bobo$2b2o4bo3bo4b2o$4bo4b3o4bo$2bo2bo2b2o
b2o2bo2bo$3bo2bo7bo2bo$6b4ob4o$5bobo5bobo2$3b2o3bo3bo3b2o$3bobo2b2ob2o
2bobo$5b3o5b3o$2b2o2bo2bobo2bo2b2o$6bo7bo$7b7o$5bobo5bobo$4b2obo5bob2o
$10bo$6bo3bo3bo$4b2o4bo4b2o2$4b2o2b2ob2o2b2o$4b2o2b2ob2o2b2o$4bobobobo
bobobo$2b2ob2o7b2ob2o$2bo2bo9bo2bo$bobo3b7o3bobo$6b9o2$7bob3obo$8bobob
o$5b2obo3bob2o$4bobo7bobo$3b2obob2ob2obob2o$2bobobo7bobobo$b2o3bobo3bo
bo3b2o$o3bo4bobo4bo3bo$3bo2b2o5b2o2bo$2o8bo8b2o$9b3o$4bo3b5o3bo$3b3ob
2o3b2ob3o$2b2o3bo5bo3b2o$b2o2bob7obo2b2o$2bo4bobobobo4bo$bobo13bobo$o
3b3obobobob3o3bo!
What do you do with ill crystallographers? Take them to the mono-clinic!

User avatar
LaundryPizza03
Posts: 1061
Joined: December 15th, 2017, 12:05 am
Location: Unidentified location "https://en.wikipedia.org/wiki/Texas"

Re: Catagolue Oddities

Post by LaundryPizza03 » October 27th, 2020, 12:02 am

Catagolue no longer animates preview images for certain periodic objects with period below 100, but the images are still fit to the largest phase (by bounding box), so patterns with large variations in bounding box, such as the object on this census page, have awkwardly cropped preview stills.

Code: Select all

x = 4, y = 3, rule = B3-q4z5y/S234k5j
2b2o$b2o$2o!
LaundryPizza03 at Wikipedia

The latest edition of new-gliders.db.txt and oscillators.db.txt have 31150 spaceships and 1205 oscillators from outer-totalistic rules. You are invited to help!

Post Reply