Catagolue Discussion Thread

[GUYTU6J]

Random question: there are a total of 45919 entries in the synthesis-costs/xp2 tabulation, the first 1168 of which are true objects. Among the remaining pseudo objects, how many are not some intermediate stage of other synthes[i/e]s and what are they?

[dvgrn]

Random question: there are a total of 45919 entries in the synthesis-costs/xp2 tabulation, the first 1168 of which are true objects. Among the remaining pseudo objects, how many are not some intermediate stage of other synthes[i/e]s and what are they?

Most of the rest aren't actually pseudo objects, they're just constellations, or at best constellations containing some pseudo objects. So I'd like to add to this question:

how many of those not-intermediate-stage syntheses are really syntheses that someone meant to add?
which ones are accidents caused by Catagolue misinterpreting a stamp-collection synthesis and constructing some silly dead-end random constellation?
what's the best way to get rid of these junk syntheses?

[LaundryPizza03]

The number of INT rules in Catagolue just surpassed 10000. I'm not sure which rule was the 10,000th to receive a haul. My best guess is that I did it with b2cek3a4a5cekqy6cs1c2-cn3eir4eiry5e6cn7c.

[AlbertArmStain]

I think the Catagolue should have the ability to identify and name still lifes by Composite patterns as well as APGcode

For example:
xs15_3lkia4z32 would change to Eater-wield Snorkel loop

And xs41_8e1eo0kcwo4oz33034al913kp would change to block on hat siamese barge with long tail bridge snake siamese short Z weld python bridge beehive

[yujh]

I think the Catagolue should have the ability to identify and name still lifes by Composite patterns as well as APGcode

For example:
xs15_3lkia4z32 would change to Eater-wield Snorkel loop

And xs41_8e1eo0kcwo4oz33034al913kp would change to block on hat siamese barge with long tail bridge snake siamese short Z weld python bridge beehive

It is able to identify object names if they are named

[AlbertArmStain]

I think the Catagolue should have the ability to identify and name still lifes by Composite patterns as well as APGcode

For example:
xs15_3lkia4z32 would change to Eater-wield Snorkel loop

And xs41_8e1eo0kcwo4oz33034al913kp would change to block on hat siamese barge with long tail bridge snake siamese short Z weld python bridge beehive

It is able to identify object names if they are named

I’m talking about unnamed objects.

[yujh]

I think the Catagolue should have the ability to identify and name still lifes by Composite patterns as well as APGcode

For example:
xs15_3lkia4z32 would change to Eater-wield Snorkel loop

And xs41_8e1eo0kcwo4oz33034al913kp would change to block on hat siamese barge with long tail bridge snake siamese short Z weld python bridge beehive

It is able to identify object names if they are named

I’m talking about unnamed objects.

So they do not have names, most likely they do not need names

[hkoenig]

I think the Catagolue should have the ability to identify and name still lifes by Composite patterns as well as APGcode

Are you volunteering to do a demonstration project doing this? Because something like this is going to be non-trivial, and require a lot of preparation. But then end result should be interesting.

The first part of the project would be to set up and distributea database of common parts and inductors that would be useful. A while back, someone presented the idea of "Life islands", and I hacked up my editor to support the building of objects based on them. Then spent some time by hand coming up with smaller Islands to add to the catalog. (Islands are simply-connected stable regions, with zero or more birth cells that need to be suppressed. Lots of them are Inductors. All Stable objects consist of one or more Islands.) The catalog is by no means comprehensive, but does have 250 entries, consisting of many (most?) islands of 10 bits or less, and examples up to 16 bits. Next step would probably be to write a program to exhaustively find all these islands, as various people have done searching for Stable Objects or Period 2 Oscillators.

Next you might want to also add in definitions for all those terms you want to use -- "block on hat siamese barge with long tail bridge snake siamese short Z weld python bridge beehive" -- and how they get applied. Some of them aren't really Islands, and so you'd need to also specify how they can legally connect to each other.

Then you need to figure out how you prioritize which name is used when your name generator algorithm comes up with more than one.

I haven't had the time to work on "Life Islands" since then, but could send you my catalog list to you get started.

So they do not have names, most likely they do not need names

I tend to agree. An objective designation like an apgcode works just fine to identify most objects.

[dani]

I think the Catagolue should have the ability to identify and name still lifes by Composite patterns as well as APGcode

For example:
xs15_3lkia4z32 would change to Eater-wield Snorkel loop

And xs41_8e1eo0kcwo4oz33034al913kp would change to block on hat siamese barge with long tail bridge snake siamese short Z weld python bridge beehive

This is extremely hard to do with an algorithm and is why there is no script that can do this for you. Still lifes with a lot of components that can be rearranged, like these, are near impossible to name unambiguously:

Code: Select all

x = 179, y = 39, rule = B3/S23
20bo2bo7bo2bo7bo2bo5bo2bo$20b4o7b4o7b4o5b4o2$20b2o11b2o7b2o9b2o$20bo
12bo8bo10bo$2bo2bob2o6bo2bobo7bo2bobo3bo2bobo5bo2bobo$2b4obo7b4ob2o6b
4ob2o2b4ob2o4b4ob2o7b2o12b2o12b2o7b2o$7bo54bo13bo13bo8bo$2b2o3b2o6b2o
11b2o7b2o9b2o9b2obo10b2obo10b2obo5b2obo$2b2o11b2o11b2o7b2o9b2o9b2ob2o
3b2o4b2ob2o9b2ob2o4b2ob2o$7b4o56bo$4o3bo2bo2b4o9b4o7b4o7b4o5b4ob4obo3b
4ob4ob2o2b4ob2o2b4ob2o$o2bo9bo2bo9bo2bo7bo2bo7bo2bo5bo2bobo2bob2o2bo2b
obo2bobo3bo2bobo3bo2bobo$81bo8bo8bo$81b2o7b2o7b2o2$88b4o7b4o$88bo2bo7b
o2bo3$173b2o$172bo2bo$172bo2bo$b2o4b2o7b2o4b2o4b2o4b2o16b2o10b2o5b2o
13b2o16b2o16b2o4b2o13b2o13b2o13b2ob2o$o2bo2bo2bo5bo2bo2bo2bo2bo2bo2bo
2bo14bo2bo8bo2bo3bo2bo11bo2bo14bo2bo14bo2bo2bo2bo11bo2bo11bo2bo14bo2bo
$o2bo2bo2bo5bo2bo2bo2bo2bo2bo2bo2bo14bo2bo8bo2bo3bo2bo11bo2bo14bo2bo
14bo2bo2bo2bo11bo2bo11bo2bo14bo2bo$b2ob2ob2ob2o4b2ob2ob2ob2ob2o4b2ob2o
4b2o7b2ob2o4b2ob2o5b2ob2o4b2o4b2ob2o13b2ob2o13b2ob2ob2o4b2o4b2ob2o4b2o
4b2ob2ob2o4b2o4b2ob2o$3bo2bo2bo2bo5bo2bo2bo2bo8bo2bo2bo2bo8bo2bo2bo2bo
9bo2bo2bo2bo5bo2bo14bo2bo14bo2bo5bo2bo2bo2bo5bo2bo2bo2bo2bo2bo2bo2bo2b
o2bo$3bo2bo2bo2bo5bo2bo2bo2bo8bo2bo2bo2bo8bo2bo2bo2bo9bo2bo2bo2bo5bo2b
o14bo2bo14bo2bo5bo2bo2bo2bo5bo2bo2bo2bo2bo2bo2bo2bo2bo2bo$4b2ob2ob2o7b
2o4b2o10b2ob2ob2ob2o7b2ob2ob2o11b2ob2ob2o7b2ob2o4b2o7b2ob2o10b2ob2ob2o
4b2ob2ob2ob2o4b2ob2ob2o4b2o4b2ob2ob2o$6bo2bo29bo2bo2bo2bo8bo2bo15bo2bo
11bo2bo2bo2bo8bo2bo8bo2bo2bo2bo5bo2bo2bo2bo5bo2bo14bo2bo$6bo2bo29bo2bo
2bo2bo8bo2bo15bo2bo11bo2bo2bo2bo8bo2bo8bo2bo2bo2bo5bo2bo2bo2bo5bo2bo
14bo2bo$7b2o31b2o4b2o10b2o17b2ob2o10b2ob2ob2o10b2ob2o7b2o4b2o7b2o4b2o
7b2o16b2o$79bo2bo11bo2bo14bo2bo$79bo2bo11bo2bo14bo2bo$80b2o13b2o16b2ob
2o$115bo2bo$115bo2bo$116b2o!

apgcodes seem to work just fine, and even sometimes show you what components they have - if an apgcode has something like '39c' or 'c93' in them you can bet it has a cap (edited from table because I made a brain error)

[AlbertArmStain]

I think the Catagolue should have the ability to identify and name still lifes by Composite patterns as well as APGcode

For example:
xs15_3lkia4z32 would change to Eater-wield Snorkel loop

And xs41_8e1eo0kcwo4oz33034al913kp would change to block on hat siamese barge with long tail bridge snake siamese short Z weld python bridge beehive

This is extremely hard to do with an algorithm and is why there is no script that can do this for you. Still lifes with a lot of components that can be rearranged, like these, are near impossible to name unambiguously:

Code: Select all

x = 179, y = 39, rule = B3/S23
20bo2bo7bo2bo7bo2bo5bo2bo$20b4o7b4o7b4o5b4o2$20b2o11b2o7b2o9b2o$20bo
12bo8bo10bo$2bo2bob2o6bo2bobo7bo2bobo3bo2bobo5bo2bobo$2b4obo7b4ob2o6b
4ob2o2b4ob2o4b4ob2o7b2o12b2o12b2o7b2o$7bo54bo13bo13bo8bo$2b2o3b2o6b2o
11b2o7b2o9b2o9b2obo10b2obo10b2obo5b2obo$2b2o11b2o11b2o7b2o9b2o9b2ob2o
3b2o4b2ob2o9b2ob2o4b2ob2o$7b4o56bo$4o3bo2bo2b4o9b4o7b4o7b4o5b4ob4obo3b
4ob4ob2o2b4ob2o2b4ob2o$o2bo9bo2bo9bo2bo7bo2bo7bo2bo5bo2bobo2bob2o2bo2b
obo2bobo3bo2bobo3bo2bobo$81bo8bo8bo$81b2o7b2o7b2o2$88b4o7b4o$88bo2bo7b
o2bo3$173b2o$172bo2bo$172bo2bo$b2o4b2o7b2o4b2o4b2o4b2o16b2o10b2o5b2o
13b2o16b2o16b2o4b2o13b2o13b2o13b2ob2o$o2bo2bo2bo5bo2bo2bo2bo2bo2bo2bo
2bo14bo2bo8bo2bo3bo2bo11bo2bo14bo2bo14bo2bo2bo2bo11bo2bo11bo2bo14bo2bo
$o2bo2bo2bo5bo2bo2bo2bo2bo2bo2bo2bo14bo2bo8bo2bo3bo2bo11bo2bo14bo2bo
14bo2bo2bo2bo11bo2bo11bo2bo14bo2bo$b2ob2ob2ob2o4b2ob2ob2ob2ob2o4b2ob2o
4b2o7b2ob2o4b2ob2o5b2ob2o4b2o4b2ob2o13b2ob2o13b2ob2ob2o4b2o4b2ob2o4b2o
4b2ob2ob2o4b2o4b2ob2o$3bo2bo2bo2bo5bo2bo2bo2bo8bo2bo2bo2bo8bo2bo2bo2bo
9bo2bo2bo2bo5bo2bo14bo2bo14bo2bo5bo2bo2bo2bo5bo2bo2bo2bo2bo2bo2bo2bo2b
o2bo$3bo2bo2bo2bo5bo2bo2bo2bo8bo2bo2bo2bo8bo2bo2bo2bo9bo2bo2bo2bo5bo2b
o14bo2bo14bo2bo5bo2bo2bo2bo5bo2bo2bo2bo2bo2bo2bo2bo2bo2bo$4b2ob2ob2o7b
2o4b2o10b2ob2ob2ob2o7b2ob2ob2o11b2ob2ob2o7b2ob2o4b2o7b2ob2o10b2ob2ob2o
4b2ob2ob2ob2o4b2ob2ob2o4b2o4b2ob2ob2o$6bo2bo29bo2bo2bo2bo8bo2bo15bo2bo
11bo2bo2bo2bo8bo2bo8bo2bo2bo2bo5bo2bo2bo2bo5bo2bo14bo2bo$6bo2bo29bo2bo
2bo2bo8bo2bo15bo2bo11bo2bo2bo2bo8bo2bo8bo2bo2bo2bo5bo2bo2bo2bo5bo2bo
14bo2bo$7b2o31b2o4b2o10b2o17b2ob2o10b2ob2ob2o10b2ob2o7b2o4b2o7b2o4b2o
7b2o16b2o$79bo2bo11bo2bo14bo2bo$79bo2bo11bo2bo14bo2bo$80b2o13b2o16b2ob
2o$115bo2bo$115bo2bo$116b2o!

apgcodes seem to work just fine, and even sometimes show you what components they have - if an apgcode has something like '39c' or 'c93' in them you can bet it has a table

Oh, I forgot about the rearrangement.

I would like the find by agpcode feature back while still having the identification by RLE feature. I remember when you could type in an agpcode and the object would comeup.

[hotdogPi]

I would like the find by agpcode feature back while still having the identification by RLE feature. I remember when you could type in an agpcode and the object would comeup.

You can still type it into the URL.

[Book]

I would like the find by agpcode feature back while still having the identification by RLE feature. I remember when you could type in an agpcode and the object would comeup.

You can still type it into the URL.

Or use Pattern Lookup: https://golhobby.com/

[dvgrn]

I've started running the auto-upload script and doing comparisons between glider synth costs in Catagolue and the LifeWiki.

People have been doing really well at keeping things up to date: so far there has been only one case where the LifeWiki synth cost was higher than Catagolue's (62P7, now fixed) and one case where Catagolue's synth cost was higher than the LifeWiki's (Bi-gun) -- 10 gliders instead of Catagolue's 14.

Code: Select all

x = 82, y = 47, rule = B3/S23
59bobo$59b2o$60bo3$obo$b2o$bo11$47bobo$48b2o$48bo$32b2o$27bo4bobo18bob
o$27b2o3bo16bo3b2o$26bobo18bobo4bo$48b2o$33bo$32b2o$32bobo11$80bo$79b
2o$79bobo3$21bo$21b2o$20bobo!

I know a little more about Shinjuku than I used to, but it's still a minor mystery to me exactly how to successfully sneak a synthesis of an infinite-growth pattern into Catagolue. Anyone feel like patching up the discrepancy, preferably in walkthrough form to explain exactly how it's done?

[Ian07]

I know a little more about Shinjuku than I used to, but it's still a minor mystery to me exactly how to successfully sneak a synthesis of an infinite-growth pattern into Catagolue. Anyone feel like patching up the discrepancy, preferably in walkthrough form to explain exactly how it's done?

Right now, the least inconvenient way to do this is by editing initialise/diffupdate.py, which APG has just done. Note the warning at the beginning:

Code: Select all

# RLE syntheses for linear-growth patterns (update these, but be careful
# that they're absolutely correct before committing because there are no
# validity checks):
always_include = '''
#CSYNTH yl78_1_20_ce2bb2c697ad330277935c32ef179035 costs 14 gliders (true).
x = 74, y = 80, rule = B3/S23
66bobo$66b2o$67bo14$39bo11bobo$39bobo9b2o$39b2o11bo3$54b3o$54bo$55bo7$
51bo$49b2o$50b2o4$23bo$24bo$22b3o46b3o$bo69bo$2bo69bo$3o46b3o$49bo$50b
o4$22b2o$23b2o$22bo7$18bo$19bo$17b3o3$21bo11b2o$21b2o9bobo$20bobo11bo
14$6bo$6b2o$5bobo!

#CSYNTH yl1344_2_1118_c2aaaf958f8f412c06554ab3e40b8749 costs 6 gliders (true).
x = 23, y = 23, rule = B3/S23
9bobo$10b2o$10bo6bo$16bo$16b3o$20b2o$20bobo$20bo2$o$b2o$2o5$3b2o$2bob
o$4bo2$5b3o$5bo$6bo!

#CSYNTH yl46_1_10_ace82e6a60e972a3a224b998d5e29b17 costs 10 gliders (true).
x = 82, y = 47, rule = B3/S23
59bobo$59b2o$60bo3$obo$b2o$bo11$47bobo$48b2o$48bo$32b2o$27bo4bobo18bob
o$27b2o3bo16bo3b2o$26bobo18bobo4bo$48b2o$33bo$32b2o$32bobo11$80bo$79b
2o$79bobo3$21bo$21b2o$20bobo!
'''

Annoyingly, since this method was only established in August 2020, only the most recently added syntheses are included in diffupdate.py. As for the original six, (i.e. pufferfish, GGG, original bi-gun, AK-94, BLSE, GPSE) I have no idea how they were originally added.

P.S. I should note that this discrepancy had existed for quite a while - since October 2019. In hindsight I probably should've left the "apgcode" infobox parameter alone, as this would've gotten fixed a lot sooner. I finally reinstated it in September 2021, and I guess this is the first time the auto-upload script has been run since then.

[confocaloid]

On pages for spaceships, shouldn't LifeViewer automatically use TRACK feature? For example, something like this:

Code: Select all

x = 9, y = 9, rule = B3-k/S23
2o2b2o$2o3bob2o$bo3bobo$6bo$bo$b2o$o2bo$obo$bo!
#C [[ GRID THUMBLAUNCH THUMBSIZE 2 THEME Catagolue ]]
#C [[ TRACK 0/7 1/7 ]]

on this page.

Another question/suggestion: On pages like this, should "All-time contribution leaderboard" and/or "Monthly contribution leaderboard" be column charts (perhaps with logarithmic scale y-axis) rather than circle charts? I thought about this and it seems to me that this would make relatively small contributions still visible even when another user contributed orders of magnitude more objects. Would logscale bar charts be an improvement over circle charts?

[yujh]

I will be against the track feature for some reasons. I will also dislike the second for some obvious reasons. The pie chart clearly demonstrates the percentage of a user’s contribution; it is not that hard to get yours visible and only takes 1 year for you name to show up in case your computer is old(I think, won’t be much longer)

[hotdogPi]

I will be against the track feature for some reasons. I will also dislike the second for some obvious reasons. The pie chart clearly demonstrates the percentage of a user’s contribution; it is not that hard to get yours visible and only takes 1 year for you name to show up in case your computer is old(I think, won’t be much longer)

One year sounds like a long time. I've been running my search program for almost a year (started late March or early April last year), and it doubles battery use. From what I've heard, apgsearch is even harder on computers than my search program, and this is especially true if it's constantly running, which mine isn't.

[yujh]

I will be against the track feature for some reasons. I will also dislike the second for some obvious reasons. The pie chart clearly demonstrates the percentage of a user’s contribution; it is not that hard to get yours visible and only takes 1 year for you name to show up in case your computer is old(I think, won’t be much longer)

One year sounds like a long time. I've been running my search program for almost a year (started late March or early April last year), and it doubles battery use. From what I've heard, apgsearch is even harder on computers than my search program, and this is especially true if it's constantly running, which mine isn't.

I think an average computer takes 3 months. Anyways I would rather like to see a part of the pie chart that leads to another one that has the other contributors that had less contributions

[confocaloid]

I will be against the track feature for some reasons.

Can you provide some reasons why having TRACK feature for spaceship pages is problematic? I think it would make it easier to watch the behaviour of a spaceship, and (if I'm not wrong here) it would not change the behaviour of the loaded page unless and until the viewer clicks "Launch".

I will also dislike the second for some obvious reasons. The pie chart clearly demonstrates the percentage of a user’s contribution; it is not that hard to get yours visible and only takes 1 year for you name to show up in case your computer is old(I think, won’t be much longer)

Well, it is not obvious to me. When suggesting log-scale column charts, I was not thinking about a new user who would like to apgsearch an already well-searched rule and quickly get their contributions shown. Instead, I was mainly thinking about those users who in the past contributed to a less-studied and less-apgsearched rule, and who may or may not be still active. When other users come to continue apgsearch the same rule (perhaps having faster computers and better knowledge about the rule under study), I personally believe that it would be nice if the contributions of the first users who apgsearched that rule still remained visible even after new users contributed 100 or 1000 times more objects. That said, I do understand that opinions may differ - that's why I'm posting in the discussion thread rather than suggestions thread.

Anyways I would rather like to see a part of the pie chart that leads to another one that has the other contributors that had less contributions

(If I correctly understand your suggestion) I believe it is preferable to have a single chart for all shown contributions. Having a separate sub-chart seems problematic because it requires a more-or-less artificial distinction (an arbitrary threshold) between those users whose contributions show on the main chart and those users whose contributions are lumped together on the main chart and are deferred to the sub-chart.

What about having several available ways to display the same data (so that each viewer can choose either a pie chart, a bar chart, or possibly another visualisation)?

[yujh]

I will be against the track feature for some reasons.

Can you provide some reasons why having TRACK feature for spaceship pages is problematic? I think it would make it easier to watch the behaviour of a spaceship, and (if I'm not wrong here) it would not change the behaviour of the loaded page unless and until the viewer clicks "Launch".

it differs between different people, i do not like the track feature.(it makes everything look stationary)

log scale column charts might make things very confusing.

[ihatecorderships]

it differs between different people, i do not like the track feature.(it makes everything look stationary)

log scale column charts might make things very confusing.

I do not like it either, the moving grid just well... looks weird. Also the boundary from 1 gen to another looks "skewed", I guess.
I think the pie charts are fine right now as they are, and logarithmic bar charts would make it more confusing.

[LaundryPizza03]

I can't seem to run B0 rules anymore. I may have encountered this issue before.

Code: Select all

> ./apgluxe --rule b01s025 -n 10000 --symmetry C1 -v 0
apgluxe v5.32-ll2.4.14: Rule b2kn3aeiry4kz5j6kn8s2-ci3-ac4aijwyz5ajr6-i7c8 does not match desired rule b01s025.
sh -c 'rm -f apgluxe *.o */*.o *.op */*.op *.gdca */*.gcda *.profraw *.profdata || true'
echo Clean done
Clean done
Ensuring lifelib is up-to-date...
Configuring rule b01s025; symmetry C1
Using /Library/Frameworks/Python.framework/Versions/3.9/bin/python3 to configure lifelib...
Valid symmetry: C1
Success!
g++ -c -Wall -Wextra -pedantic -O3 -pthread -flto -march=native --std=c++11  main.cpp -o main.o
g++ -c -Wall -Wextra -pedantic -O3 -pthread -flto -march=native --std=c++11  includes/md5.cpp -o includes/md5.o
g++ -c -Wall -Wextra -pedantic -O3 -pthread -flto -march=native --std=c++11  includes/happyhttp.cpp -o includes/happyhttp.o
g++ -pthread -flto  main.o includes/md5.o includes/happyhttp.o    -o apgluxe
true
true                                                oo o
true                                                oo ooo
true                                                      o
true                                                oo ooo
true                                                 o o
true                                                 o o
true                                                  o
apgluxe v5.32-ll2.4.14: Rule b01s025 is correctly configured.
apgluxe v5.32-ll2.4.14: Symmetry C1 is correctly configured.

Greetings, this is apgluxe v5.32-ll2.4.14, configured for b01s025/C1.

Lifelib version: ll2.4.14
Compiler version: Apple LLVM 13.0.0 (clang-1300.0.29.3)
Python version: '3.9.0 (v3.9.0:9cf6752276, Oct  5 2020, 11:29:23)  [Clang 6.0 (clang-600.0.57)]'

Using seed k_SWqVqqBWtYjR
Instruction set AVX2 detected
Running 10000 soups per haul:
Rule b01s025 cannot be iterated 1 x 2^0 generations.

[yujh]

I can't seem to run B0 rules anymore. I may have encountered this issue before.

Code: Select all

> ./apgluxe --rule b01s025 -n 10000 --symmetry C1 -v 0
apgluxe v5.32-ll2.4.14: Rule b2kn3aeiry4kz5j6kn8s2-ci3-ac4aijwyz5ajr6-i7c8 does not match desired rule b01s025.
sh -c 'rm -f apgluxe *.o */*.o *.op */*.op *.gdca */*.gcda *.profraw *.profdata || true'
echo Clean done
Clean done
Ensuring lifelib is up-to-date...
Configuring rule b01s025; symmetry C1
Using /Library/Frameworks/Python.framework/Versions/3.9/bin/python3 to configure lifelib...
Valid symmetry: C1
Success!
g++ -c -Wall -Wextra -pedantic -O3 -pthread -flto -march=native --std=c++11  main.cpp -o main.o
g++ -c -Wall -Wextra -pedantic -O3 -pthread -flto -march=native --std=c++11  includes/md5.cpp -o includes/md5.o
g++ -c -Wall -Wextra -pedantic -O3 -pthread -flto -march=native --std=c++11  includes/happyhttp.cpp -o includes/happyhttp.o
g++ -pthread -flto  main.o includes/md5.o includes/happyhttp.o    -o apgluxe
true
true                                                oo o
true                                                oo ooo
true                                                      o
true                                                oo ooo
true                                                 o o
true                                                 o o
true                                                  o
apgluxe v5.32-ll2.4.14: Rule b01s025 is correctly configured.
apgluxe v5.32-ll2.4.14: Symmetry C1 is correctly configured.

Greetings, this is apgluxe v5.32-ll2.4.14, configured for b01s025/C1.

Lifelib version: ll2.4.14
Compiler version: Apple LLVM 13.0.0 (clang-1300.0.29.3)
Python version: '3.9.0 (v3.9.0:9cf6752276, Oct  5 2020, 11:29:23)  [Clang 6.0 (clang-600.0.57)]'

Using seed k_SWqVqqBWtYjR
Instruction set AVX2 detected
Running 10000 soups per haul:
Rule b01s025 cannot be iterated 1 x 2^0 generations.

Interesting how I just asked this question in the discord less than two week ago.
Use

Code: Select all

git checkout 86f8c8

[hotdogPi]

C1 smiley: 90
C1 Achim's p8: 45 (exactly half as common)

G1 smiley: 214
G1 Achim's p8: 131 (would by 107 if it was half as common like C1)

Is this outside the margin of error?

[LaundryPizza03]

C1 smiley: 90
C1 Achim's p8: 45 (exactly half as common)

G1 smiley: 214
G1 Achim's p8: 131 (would by 107 if it was half as common like C1)

Is this outside the margin of error?

We can use the chi-squared test for this question.

Occurrences per soup:
Smiley: 1.04883e-12 (C1), 9.10517e-13 (G1)
Achim's p8: 5.24416e-13 (C1), 5.57373e-13 (G1)

chi^2 for smiley is (214 - 246.508)^2/246.508 = 4.28703; p-value about 0.0384041 for 1 dof
chi^2 for Achim's p8 is (131 - 123.254)^2/123.254 = 0.486784; p-value about 0.485365 for 1 dof

So smiley is significantly more common in G1 at the 5% significance level, but Achim's p8 is not.