ConwayLife.com

calcyman wrote: ↑

December 30th, 2020, 7:59 am

go to the object page, then replace 'object' with 'attribute', then append '/ikpx2_stdin' to the URL

That works fine for the c/4 ships, as in the example you gave, but for the c/2 ships such as this one, it only seems to return the apgcode.

Sokwe wrote: ↑

December 30th, 2020, 3:27 pm

That works fine for the c/4 ships, as in the example you gave, but for the c/2 ships such as this one, it only seems to return the apgcode.

Those hauls have been deleted.

I really want a way to see the soup/object/haul count in raw data form like /textcensus/.

Is there no way to recover the information on discoverer and discovery date? And is there any way to ensure that this data is archived for future discoveries? I would like to add the patterns to they moving-objects collection, and any discovery information would be useful.

Sokwe wrote: ↑

December 31st, 2020, 10:31 pm

Is there no way to recover the information on discoverer and discovery date? And is there any way to ensure that this data is archived for future discoveries? I would like to add the patterns to they moving-objects collection, and any discovery information would be useful.

No, there is no way to retrieve information from deleted hauls, but you can preserve the hauls by calling /attribute before they are deleted.

The higher-symmetry censuses for CGOL haven't been contributed to in months. Can someone contribute to them? There is still potential for new discoveries.

Also, can anyone precompile D8_1 or D8_4(or others)(on windows 10?)

yujh wrote: ↑

January 8th, 2021, 4:47 am

Also, can anyone precompile D8_1 or D8_4(or others)(on windows 10?)

I made some builds for a few higher symmetries - available here: https://conwaylife.com/forums/viewtopic ... 44#p118544

We should probably do something to encourage users to search for higher symmetries, but what can we do?

Still lives, even large and rare ones (!) are discarded in G1 soups. Large and rare still lives should be added to the census.

creeperman7002 wrote: ↑

January 11th, 2021, 10:32 am

Still lives, even large and rare ones (!) are discarded in G1 soups. Large and rare still lives should be added to the census.

How would you efficiently determine whether the ash of a soup contains a large/rare still-life?

calcyman wrote:How would you efficiently determine whether the ash of a soup contains a large/rare still-life?

How is it currently done? I'd legitimately like to know. I've been rewriting a bunch of my Life tools, and one of the features I've lost is the ability to detect escaping Gliders. (The old way was arbitrary and inefficient and not compatible with the new methods.)

Identifying objects is a little improved, as I've switched to using modified apgcodes, which makes searching the database a bit easier. But it sometimes breaks down when I try to partition pseudo-objects into their constituent components, or have to deal with long period oscillators.

(Long time ago I implemented most Matthew Cook's "Still Life Theory" up to the ambiguous 3 or 4 object partitioning. It seemed to work, or at least passed all my test cases. I looked at that C++ code recently and decided the implementor did a poor job that can't or shouldn't be salvaged.)

So where in the apgsearch code and such are the object identification and partitioning algorithms? What are the criteria for saying that stabilization has occurred? How are the infinite objects (guns and puffers) identified? Mostly I'd just like some hints as to where to start looking for ideas and algorithms to borrow.

creeperman7002 wrote: ↑

January 11th, 2021, 10:32 am

Still lives, even large and rare ones (!) are discarded in G1 soups. Large and rare still lives should be added to the census.

And p2 and p3s. How do you propose to identify them? (Without scanning over every part of the soup; at that point you should just add everything and call it C1.)

Can there be a tabulation of the most recently discovered objects along with their discovery date and discoverer?

It would seem that we have reached 1 quadrillion objects in B3/S23/C1.

LaundryPizza03 wrote: ↑

January 25th, 2021, 10:20 am

It would seem that we have reached 1 quadrillion objects in B3/S23/C1.

According to wwei, this is the haul containing the quadrillionth object.

A for awesome wrote: ↑

January 25th, 2021, 10:25 am

According to wwei, this is the haul containing the quadrillionth object.

Actually, dani said that in #announcements. I just saved it in the Wayback Machine.

The 100 most recent hauls in B3/S23/C1 produced a total of 111016381538 objects from 5081798813 soups (average 21.84593153 objects/soup). The weighted sample variance among these hauls was 0.0045303(29) soups/object. If we extrapolate this to all 1801630 hauls (total of 46413097249131 soups), the total uncertainty is approximately 4.74048(31) * 10^-6.

A total of 1013891146973773 objects were found from 46413097249131 soups, for an average of

n = 21.8449362(47) objects/soup.

We can also compute by scaling the standard deviation backwards to the individual soup. By doing so, we find

Δn = 32.458(21) objects/soup.

Also, Catagolue's wiki page needs to be updated.

Saka wrote: ↑

August 23rd, 2017, 7:20 am

A thread for discussing about Catagolue, but not apgsearch. Muzik requested this because the apgsearch threads were getting a bit cluttered with Catagolue discussion, so I thought I should make one of these "General Discussion" threads.

Now,
Is Catagolue ever going to get animated generations, b0, and LTL pattern svgs?
Are we ever going to be able to customize our apgprofiles?

Requested Edit By Danny:
Catagolue resets every 7:00 UTC

Sorry, I can't login in on Hαtsyα Catagolue. It leads me to a Google page.
Also, the BLSE synthesis given on Catagolue has two extra beehives and an extra block

There is a typo in the original post.

First of all: my first post here as a beginning Game of Life enthusiast, pleasure to meet you all

Out of pure interest in what I'm able to find in all those random soups I have been contributing some hauls to Catagolue over the last several days using wildmyron's apgsearch builds in this thread. Tried the C1, 1x256 and D8_4 symmetries and everything uploads fine, but the D8_4 hauls I submitted don't seem to appear in the hauls list. I did find my and several other submitted hauls on this page though, so perhaps there might be something going on with the validation process. Are newly submitted hauls for the D8_4 symmetry only verified sporadically, of is there some sort of issue with the apgsearch build I'm using?

Thanks in advance for answering, and onto more cool finds!

GoldfishBert wrote: ↑

February 13th, 2021, 8:39 am

First of all: my first post here as a beginning Game of Life enthusiast, pleasure to meet you all

Out of pure interest in what I'm able to find in all those random soups I have been contributing some hauls to Catagolue over the last several days using wildmyron's apgsearch builds in this thread. Tried the C1, 1x256 and D8_4 symmetries and everything uploads fine, but the D8_4 hauls I submitted don't seem to appear in the hauls list. I did find my and several other submitted hauls on this page though, so perhaps there might be something going on with the validation process. Are newly submitted hauls for the D8_4 symmetry only verified sporadically, of is there some sort of issue with the apgsearch build I'm using?

Thanks in advance for answering, and onto more cool finds!

Welcome to the forum, I hope you find some interesting ideas here and some challenges to pursue.

I'm not sure which hauls specifically are yours but I presume they're the latest which were submitted in the last few days. It is indeed the case that there is an issue with the verification process, but it is actually by design. Catagolue tries to prevent hauls which are not authentic, or tampered with in some way, from being committed to the census (for all the official symmetries in CGoL and other popular symmetries in Life-like rules). Part of this is to prevent verification by anonymous users and self-verification when the backlog of unverified hauls is less than 100 hauls[*]. For symmetries where only one user is contributing this means there's generally a large backlog that doesn't go away. Unfortunately, a lot of the higher symmetries haven't been popular lately, hence your hauls haven't been verified.

However, I don't see any hauls that have been verified by you which is surprising. I'm not sure why that would be unless you disabled verification or there is an issue with verification in those builds you're using. (Any such issue wouldn't have anything to do with why your builds haven't been verified though).

Occasionally I do a verification run in which I instruct apgluxe to verify a large number of hauls without actually committing anything itself for all the higher symmetries in CGoL. I've kicked that off a little while ago so you should see your hauls in the census a bit later today. Unfortunately that won't do anything for the hauls you submit into the future, but they will make it into the census eventually.

[*] Personally I don't think preventing self-verification is an effective defense against an attack trying to get an invalid soup into the census because it's relatively easy to bypass. However, it does reduce the chance of someone accidentally submitting invalid hauls with a buggy version of apgluxe by ensuring someone else (who most likely has a different version) verifies them.

Edit: Hmm, bad news. most of those hauls ended up here: https://catagolue.hatsya.com/haul/b3s23 ... ommitted=4 which means they didn't pass. I have no idea what's going on there. Hopefully calcyman can figure out what's going on.

Edit 2: These recent hauls are now committed to the census. Calcyman has updated the chisquared threshold on Catagolue which is used to reject hauls with a statiscally abnormal number of objects per soup. Hopefully that means there won't be too many hauls incorrectly rejected in the future.

Is there any way to find a list of all objects with a certain component, e.g. a dock? https://www.google.com/search?&q=dock+s ... hatsya.com doesn't work, although it would only work for patterns that have been named anyway.

MathAndCode wrote: ↑

February 26th, 2021, 11:46 pm

Is there any way to find a list of all objects with a certain component, e.g. a dock? https://www.google.com/search?&q=dock+s ... hatsya.com doesn't work, although it would only work for patterns that have been named anyway.

No, apgcodes don't describe the pattern well enough, and plus it'd be infinite.

You probably would need to build a database-- Define a table that for each row has a column for an island's hashcode (probably apgcode) and another column for an object's apgcode. Then, for every object, it's easy to partition into the individual islands. Now add a row for each island with the associated object's apgcode. Once the database is built, you can do a query for an island by the hashcode and get a list of all objects that contain that island.

The database would be large.The time consuming part is getting the list of objects in a usable format. I've created SOF files that contain all the stable objects up to 32 bits, and that took several years. Those files take up just under half a terabyte. (The pseudo objects take up another half terabyte, too.) So I would expect the database to be larger than that, because even though apgcode is smaller than SOF, there would be duplications of the apgcode for every object with more than one island. But it should be easy to quickly create a "proof-of-concept" with objects up to 16 or 20 bits.

(Note: not using a full apgcode will make searching go faster. The part to the right of the underscore is what matters, and the info on the left is more useful if it has its own column or columns in the row. If you want object or island size, for example, add columns for that data. Also, multiple tables and SQL JOINs might work, but I prefer the simple/straitforward/brute force design for things like this.)

Is there a Catagolue-generated list of natural objects that have not been synthesized?

from statistics based on textcensus.csv




Here is 'Occurrence Summary comparation' between C1 and G1
To be more specific, C1 have 1% advantage of discovery new object.



G1 would have slightly 'little' new discovery than C1. (maybe the xp2s & xp3s)
It may suggest 'filter soups as you desire' wouldn't effect the rate of new discovery.

if increase 'filter rate' can help the search speed. I'll suggest p19_stdin dedicated searching period >=19. lol