Catagolue Oddities

[Moosey]

The Long Prodigal is called the Very Long integral for some reason
https://catagolue.appspot.com/object/xs ... aeijns234e

[Ian07]

The Long Prodigal is called the Very Long integral for some reason
https://catagolue.appspot.com/object/xs ... aeijns234e

That's from an error on the wiki which I fixed about a month ago: http://www.conwaylife.com/w/index.php?t ... ldid=60474

wiki-names.txt, however, still hasn't been updated since January.

[testitemqlstudop]

Pardon my ignorance, but why do I have Sprotsmanship? I haven't uploaded hauls to any censuses other than C1 for the last few days, and it can't be from X3VI since it has less than 333 billion objects.

http://gol.hatsya.co.uk/user/testitemqlstudop

[apg]

Pardon my ignorance, but why do I have Sprotsmanship? I haven't uploaded hauls to any censuses other than C1 for the last few days, and it can't be from X3VI since it has less than 333 billion objects.

http://gol.hatsya.co.uk/user/testitemqlstudop

Oh, there's a bug in which it believes that backup censuses (e.g. C1-2016-01-07) count as 'different symmetries'.

[Ian07]

Non-alphanumeric characters in an apgcode seem to be automatically removed when checking the pattern for periodicity and generating statistics for it, but not when creating the LifeViewer, the synthesis, the sample soups, or the comments: http://catagolue.appspot.com/object/xp3 ... 401@/b3s23

[Hdjensofjfnen]

My user page states that I've contributed at least 333,333,333,334 objects to a different rule or symmetry, but I distinctly remember having contributed only about 100 hauls of 100k soups each (Python script) to rules outside of B3/S23. Something doesn't add up, as it just appeared on my user page today and I haven't uploaded anything to non-Life rules since 2018.
EDIT: Oh, wait, I should have read testitem's comment.

[muzik]

G1 and co. still aren't classed as official symmetries. Is there a reason for this I'm not aware of?

[Ian07]

Lately the update process has been running twice each cycle: https://gitlab.com/apgoucher/catagolue/-/jobs

[Saka]

Take a look at
https://catagolue.appspot.com/census/b2 ... 4iw5y7e/C1

[testitemqlstudop]

*silent giggles*

I mean, not only do they have the same name, they have the same payosha256 keys, too...

[muzik]

For the schick engine, it shows two different c1 sample soups, but both of these link to an identical soup:

https://catagolue.appspot.com/object/xq ... xfjf/b3s23

[testitemqlstudop]

The page for the Schick Engine lists two C1 soups, but both links point to the same soup (the original one). Meanwhile, the xq12 tabulation lists only one occurrence, as expected.

Sorry for breaking the page

[muzik]

Here's some spam: https://catagolue.appspot.com/object/xp ... y011/b3s23

[mniemiec]

Here's some spam: https://catagolue.appspot.com/object/xp ... y011/b3s23

I've frequently used Catagolue to look at soups for objects that do not yet have syntheses. In many case, some of the rarer objects claim that they do have associated soups - but only have hauls from "replit_smaller_stdin_test" and "replit_stdin_test". Examining those hauls more closely shows that all the soups in there are fraudulent, and they contain a huge plethora of rare objects of various. The also have catagories xsInfo, xsWarning, ylInfo, and ylWarning - whose sole purpose is to mock apgsearch's lax security measures.

There should be a mechanism for manually purging the detabase of known fraudulent hauls.

[Ian07]

There should be a mechanism for manually purging the detabase of known fraudulent hauls.

Normally, these types of symmetries would be moved to the "Troll symmetries" section of the B3/S23 page, which prevents them from showing up as sample soups. Right now here's the code that marks these as troll symmetries:

Code: Select all

        } else if (symmetry.contains("wwei") || symmetry.contains("nonymous") || symmetry.contains("256X2")) {
            // Troll symmetry:
            symstat = 3;
        } else if (symmetry.equals("Uploads") || symmetry.equals("Saka_Test") || symmetry.equals("heavgolue")) {
            // More troll symmetries:
            symstat = 3;

I've submitted a merge request to Catagolue so that the "replit" symmetries will also be considered trolls.

[mniemiec]

Normally, these types of symmetries would be moved to the "Troll symmetries" section of the B3/S23 page, which prevents them from showing up as sample soups. Right now here's the code that marks these as troll symmetries: ... I've submitted a merge request to Catagolue so that the "replit" symmetries will also be considered trolls.

Couldn't 256X2 be a legitimate symmetry, or are they not valid random soups? The problem with the "replit" symmetries is that the soups themselves are fraudulent - they contain patterns that claim to produce certain objects, while not actually producing them. At least with an uploaded soup, someone could upload a "foo predecessor" which actually does generate a "foo" object, even if its makeup is statistically dubious. "replit" doesn't even do that.

Ideally, it should be possible to totally remove some of the more egregious troll-symmetric hauls from the database altogether.

[Ian07]

Here's another oddity: the b3s23/C1 census lists a yl1152_1_332_ce5c623ad49c2712d55df5b89c87da86, yet it doesn't have any sample soups.

[GUYTU6J]

flaw1.png (4.16 KiB) Viewed 11499 times

[muzik]

5 committed hauls?

https://catagolue.appspot.com/census/b2 ... 5ar6i7e/C1

[apg]

5 committed hauls?

https://catagolue.appspot.com/census/b2 ... 5ar6i7e/C1

That's Snowflakes, right? I had to restore the census after DefinitelyAlphanumeric deleted it, and the haul count was reset.

[Ian07]

Why are these considered pseudo-objects?
https://catagolue.appspot.com/object/xp ... w121/b3s23
https://catagolue.appspot.com/object/xp ... d1dd/b3s23
https://catagolue.appspot.com/object/xp ... zw11/b3s23
https://catagolue.appspot.com/object/xp ... l5lo/b3s23
https://catagolue.appspot.com/object/xp ... pl91/b3s23
https://catagolue.appspot.com/object/xp ... 5031/b3s23

(Note: These are just a few examples. After scrolling through the pseudo sections of the oscillator tabulations of the synthesis-costs census I noticed several more.)

[apg]

Why are these considered pseudo-objects?
https://catagolue.appspot.com/object/xp ... w121/b3s23
https://catagolue.appspot.com/object/xp ... d1dd/b3s23
https://catagolue.appspot.com/object/xp ... zw11/b3s23
https://catagolue.appspot.com/object/xp ... l5lo/b3s23
https://catagolue.appspot.com/object/xp ... pl91/b3s23
https://catagolue.appspot.com/object/xp ... 5031/b3s23

(Note: These are just a few examples. After scrolling through the pseudo sections of the oscillator tabulations of the synthesis-costs census I noticed several more.)

Oh, I made a mistake last night and will be correcting it relatively soon.

EDIT: normality restored. That should also apply to those *WSS flotillae misclassified as pseudo.

[Ian07]

This was submitted to the nonmagical box last night by Alex Greason:

Code: Select all

x = 18, y = 22, rule = B3/S23
2bo$obo$b2o2$15bo$15bobo$15b2o7$9bo$8bobo$8bobo$9bo2$13b2o$12bo2bo$13b
obo$14bo!

The update job interpreted this correctly:

Code: Select all

# Engulfing 8k4oy7654zw1w8kk8zzyf351
# processing src=xs13_696zy02596, dst=xs15_33z0oozx4aic...
#     ...success.

However, despite the initial step costing 3G, the destination still life still costs 9G.

[apg]

This was submitted to the nonmagical box last night by Alex Greason:

Code: Select all

x = 18, y = 22, rule = B3/S23
2bo$obo$b2o2$15bo$15bobo$15b2o7$9bo$8bobo$8bobo$9bo2$13b2o$12bo2bo$13b
obo$14bo!

The update job interpreted this correctly:

Code: Select all

# Engulfing 8k4oy7654zw1w8kk8zzyf351
# processing src=xs13_696zy02596, dst=xs15_33z0oozx4aic...
#     ...success.

However, despite the initial step costing 3G, the destination still life still costs 9G.

The 3G initial step isn't in Shinjuku, so it must have been deleted at some point. (I think Alex Greason's script begins by flagging any objects where the local Shinjuku working tree disagrees with Catagolue's synthesis-costs census about their costs. Potentially these syntheses could be automatically downloaded from Catagolue and fed into the nonmagical box; that would rectify these issues.)

[praosylen]

https://catagolue.appspot.com/object/xs ... 11dd/b3s23