Shinjuku: a database of glider syntheses

For general discussion about Conway's Game of Life.
User avatar
Ian07
Posts: 808
Joined: September 22nd, 2018, 8:48 am

Re: Shinjuku: a database of glider syntheses

Post by Ian07 » March 9th, 2020, 6:09 am

Shinjuku seems to be having a hard time with XWSSes all of a sudden. These RLEs, for example, was rejected as invalid, even though they should be reversible without a problem as far as I know:

Code: Select all

x = 0, y = 0, rule = B3/S23
32bo$30bobo$31b2o35bobo$26bobo39b2o$27b2o40bo$27bo11$84bo$82b2o$83b2o
8$b2o$4o44b2o$2ob2o42bobo2b2o$2b2o43bobobobo$46b2obo2bo$49bo$49b2o$9bo
41bo5b2o$10bo40bo4bobo$8b3o41b2o3bo$4b2o48bo$5b2o47b6o$4bo54bo$56b3o$
7b2o47bo$6b2o$8bo12$84b3o$84bo$85bo2$82b2o$81b2o$38bobo42bo$39b2o$39bo
8$57b2o$57bobo$57bo2$43bo$43b2o$42bobo6$58bo$57b3o$57bob2o$58b3o$58b2o
!
# Filtering with radius 5
# Engulfing yc4ocyv7420og8zyzyg1zzwac4zzyto8g0s4oyqokgy58smeczyoo8gx1o47011zyncklk4611zyq1yzy08ckzyzyha64zgggz0218ogzx1zzzyt42ey0og8zymgocoy2634w1zym133
# processing src=xs38_caab8oz0652w1168ozy26520f871, dst=xp9_oo1v0c26zx344ox8a6zy311dd113...
# Invalid component
# Filtering with radius 8
# Filtering with radius 13
# Filtering with radius 21

Code: Select all

x = 0, y = 0, rule = B3/S23
111bo$111bobo$111b2o2$110bo$110bo$110bo2$56b2o6bo36b2o3b2o$9bo45b3o6bo
bo34bobobobo$7bobo39b2o4b2obo5b2o37bobo$8b2o39b2o5b3o43bo3bo$57bo44bo
3bo$8bo54b3o37b3o$8b2o53bo$7bobo54bo38b3o$57bo44bo3bo$49b2o5b3o43bo3bo
$49b2o4b2obo44bobo$55b3o43bobobobo$56b2o43b2o3b2o2$110bo$110bo$bo108bo
$b2o$obo108b2o$111bobo$111bo!
# Filtering with radius 5
# Engulfing 1530235zyb8ck
# processing src=xs0_0, dst=xs8_33zcc...
#     ...success.
# Engulfing w33y133zcems8x8smeczy1ggz153w301
# processing src=xs8_33zcc, dst=xp2_7wggy5ggw7zxhaarle0elraahzx11y511...
# Invalid component
# Engulfing y5g742zo8g0g8ow3z069896zg6p1p6gz11x11wszy6e24
# processing src=xp2_7wggy5ggw7zxhaarle0elraahzx11y511, dst=xp2_3p606p3z0o555oz6430346...
#     ...success.
# Filtering with radius 8
# Filtering with radius 13
# Filtering with radius 21

User avatar
calcyman
Posts: 2591
Joined: June 1st, 2009, 4:32 pm

Re: Shinjuku: a database of glider syntheses

Post by calcyman » March 9th, 2020, 11:51 am

Yes, the problem is because all *WSSes are simultaneously replaced with gliders (and the resulting glider arrangement is not necessarily rewindable). If the code were modified to replace them individually, rewinding after each operation, then I believe things should work as desired.
What do you do with ill crystallographers? Take them to the mono-clinic!

User avatar
Ian07
Posts: 808
Joined: September 22nd, 2018, 8:48 am

Re: Shinjuku: a database of glider syntheses

Post by Ian07 » April 11th, 2020, 2:49 am

Question: does transfer.py try to synthesize an object if it's only used as an initial step, or does the object have to have its own complete synthesis?

Asking this because Catagolue refuses to accept the first step of this synthesis which GUYTU6J tried to submit multiple times:

Code: Select all

x = 0, y = 0, rule = B3/S23
135bo$136b2o22bo$135b2o23bobo$160b2o6$144bobo$145b2o76bo$145bo76bo$4bo
217b3o$4bobo$o3b2o$b2o217b2o$2o217bo2bo$219bo2bo$220b2o$211bo$212b2o$
211b2o2$141b2o$139b3obo$138bo4bo86bo$139b3obob2o81b3o$141bobo2bo64b2o
14bo$144b2o63b3obo13b2o$208bo4bo$209b3obob2o$211bobo2bo$214b2o5$142b3o
$144bo$143bo2$164bo61bo$163bo60b2o$163b3o59b2o$18b2o$17b2o200bobo$14b
2o3bo199b2o$13bobo204bo$15bo$216bo$216b2o$215bobo$162b2o$161bo2bobo$
161b2obob3o$164bo4bo$164bob3o63b2o$165b2o64bo2bobo$231b2obob3o$234bo4b
o$219b2o13bob3o$220bo14b2o$217b3o$217bo4$235b2o$234b2o$162bo73bo$161b
2o63b2o$161bobo61bo2bo$225bo2bo$226b2o3$223b3o$146b2o77bo$145bobo23b2o
51bo$147bo22b2o$172bo!
# Filtering with radius 5
# Engulfing 24708ok
# processing src=xs0_0, dst=xs7_178c...
#     ...success.
# Filtering with radius 8
# Filtering with radius 13
# Filtering with radius 21

User avatar
dvgrn
Moderator
Posts: 8931
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Shinjuku: a database of glider syntheses

Post by dvgrn » July 28th, 2020, 5:40 pm

Today's LifeWiki auto-upload run turned up only one case where LifeWiki knows about something that Catagolue (and therefore Shinjuku, I think?) doesn't know:
apgcodes where LifeWiki synth is better than Catagolue: [('Pre-pulsar_spaceship', 'prepulsarspaceship', 'xq30_yq232x232zug1hmgc865da808ad568cgmh1guybug1hmgc865da808ad568cgmh1guz124w6yb6w421yb124w6yb6w421', 481, 'better than', 988)]
"481" appears to be two 239-glider spider recipes plus three gliders to make a pre-pulsar. Does anyone know whether this recipe has been created, and the second spider synth actually doesn't get in the way of the first one? The last test pattern I found was from back when a spider cost 340 gliders.

There's one other item that looks like it will need a Shinjuku checkin to make Catagolue match LifeWiki stats:
apgcode where LifeWiki synth exists but no Catagolue synth: [('6-engine_Cordership', '6enginecordership', 'xq96_yk573w8aegvzyzy133zyzy9oozyao0ovogoy14440g8gzyb6ug101y61ycgggy2gw66zg8880gxg8gx4874w11yl231w4578fz34440d9ac8lya2552zyhoozy1ooggy1aghh92yc763zx77glhf2ymgmknozy5ggy9kscx8o0szy511ye11123zydcc', '37', 'not in Catagolue')]

User avatar
toroidalet
Posts: 1328
Joined: August 7th, 2016, 1:48 pm
Location: My computer
Contact:

Re: Shinjuku: a database of glider syntheses

Post by toroidalet » July 28th, 2020, 7:25 pm

The 239-glider synthesis doesn't work (one of the gliders hits the spider), but a naive hybridization (289 gliders?) works, for a total of about 531 gliders:

Code: Select all

x = 231, y = 341, rule = B3/S23
101bo$100bo$100bo$100bo2bo$100b3o48$221bo$219b2o$220b2o18$9bo$10bo$8b
3o$bo178bo$2bo177bobo$3o177b2o$15bo$16b2o$11bo3b2o160bo$12bo164bobo$
10b3o164b2o8$13bobo$14b2o$14bo170bo$143bobo39bobo$143b2o40b2o$12bo131b
o$13b2o$7bo4b2o$8b2o$7b2o$13bo$11bobo$12b2o$160bo$160bobo8bo$160b2o9bo
bo5bo$171b2o5bo$178b3o8$36b2o$33b2o2bo$33b2o2bo$38bo$33bob4o$32b2obo$
36b3o$37bo$37b2o$39bo$38bo$38bo$34b3o2$34b3o$38bo$38bo$39bo$37b2o$37bo
$36b3o$32b2obo$33bob4o$38bo$33b2o2bo$33b2o2bo$36b2o10$92bo$91bobo$92b
2o2$92b2o$92b2o6bo$87b2o7b2obobo$87bo9bobo2bo$88bo3b2o3bo2bobo$89bo3bo
4bob2o$90bo2bob3obo$91bobobo3bo$85b2o5bo3b3o$85bobo$86b2o6b5o$91bobo3b
obo$91b2ob2o3bo$94bob3o$94bobo$93b2o$94bobo$94bob3o$91b2ob2o3bo$91bobo
3bobo$86b2o6b5o$85bobo$85b2o5bo3b3o$91bobobo3bo$91bobob3obo$88bo3bo5bo
b2o$87bobo7bo2bobo$88bo8bobo2bo$91b3o2b2obobo$100bo22$164b2o$163b2o$
165bo28$14b2o$15b2o$4b2o8bo$3bobo$5bo$216b2o$216bobo$12b2o202bo$13b2o
214bo$12bo131bo35b2o46b2o$143b2o34b2o47bobo$143bobo35bo$14bo$14b2o168b
2o$13bobo167b2o$185bo7$177b2o$177bobo$154b3o20bo$15b3o136bo$17bo137bo$
16bo163b2o$180bobo$2o178bo$b2o$o$14b2o$15b2o$14bo66$100b3o$100bo2bo$
100bo$100bo$101bo!
Glide-symmetric in one more (~532):

Code: Select all

x = 231, y = 341, rule = B3/S23
101bo$100bo$100bo$100bo2bo$100b3o47$228bobo$228b2o$229bo12$203bo$203bo
bo$203b2o5$9bo$10bo$8b3o$bo178bo$2bo177bobo$3o177b2o$15bo$16b2o$11bo3b
2o160bo$12bo164bobo$10b3o164b2o8$13bobo$14b2o$14bo170bo$143bobo39bobo$
143b2o40b2o$12bo131bo$13b2o$7bo4b2o$8b2o$7b2o$13bo$11bobo$12b2o$160bo$
160bobo8bo$160b2o9bobo5bo$171b2o5bo$178b3o8$36b2o$33b2o2bo$33b2o2bo$
38bo$33bob4o$32b2obo$36b3o$37bo$37b2o$39bo$38bo$38bo$34b3o2$34b3o$38bo
$38bo$39bo$37b2o$37bo$36b3o$32b2obo$33bob4o$38bo$33b2o2bo$33b2o2bo$36b
2o10$92bo$91bobo$92b2o2$92b2o$92b2o6bo$87b2o7b2obobo$87bo9bobo2bo$88bo
3b2o3bo2bobo$89bo3bo4bob2o$90bo2bob3obo$91bobobo3bo$85b2o5bo3b3o$85bob
o$86b2o6b5o$91bobo3bobo$91b2ob2o3bo$94bob3o$94bobo$93b2o$94bobo$94bob
3o$91b2ob2o3bo$91bobo3bobo$86b2o6b5o$85bobo$85b2o5bo3b3o$91bobobo3bo$
91bobob3obo$88bo3bo5bob2o$87bobo7bo2bobo$88bo8bobo2bo$91b3o2b2obobo$
100bo22$164b2o$163b2o$165bo28$14b2o$15b2o$4b2o8bo$3bobo$5bo197b2o$202b
2o$204bo$12b2o$13b2o204b3o$12bo131bo35b2o37bo$143b2o34b2o39bo$143bobo
35bo$14bo$14b2o168b2o$13bobo167b2o$185bo7$177b2o$177bobo$154b3o20bo$
15b3o136bo$17bo137bo$16bo163b2o$180bobo$2o178bo$b2o$o$14b2o$15b2o$14bo
66$100b3o$100bo2bo$100bo$100bo$101bo!
Asymmetric in 2 more (~533):

Code: Select all

x = 236, y = 341, rule = B3/S23
101bo$100bo$100bo$100bo2bo$100b3o38$229bobo$229b2o$230bo4$227bobo$227b
2o$228bo15$7bobo$8b2o$8bo5$9bo$10bo$8b3o$bo178bo$2bo177bobo$3o177b2o$
15bo$16b2o$11bo3b2o160bo$12bo164bobo$10b3o164b2o8$13bobo$14b2o$14bo
170bo$143bobo39bobo$143b2o40b2o$12bo131bo$13b2o$7bo4b2o$8b2o$7b2o$13bo
$11bobo$12b2o$160bo$160bobo8bo$160b2o9bobo5bo$171b2o5bo$178b3o8$38b2o$
35b2o2bo$35b2o2bo$40bo$35bob4o$34b2obo$38b3o$39bo$39b2o$41bo$40bo$40bo
$36b3o2$36b3o$40bo$40bo$41bo$39b2o$39bo$38b3o$34b2obo$35bob4o$40bo$35b
2o2bo$35b2o2bo$38b2o10$92bo$91bobo$92b2o2$92b2o$92b2o6bo$87b2o7b2obobo
$87bo9bobo2bo$88bo3b2o3bo2bobo$89bo3bo4bob2o$90bo2bob3obo$91bobobo3bo$
85b2o5bo3b3o$85bobo$86b2o6b5o$91bobo3bobo$91b2ob2o3bo$94bob3o$94bobo$
93b2o$94bobo$94bob3o$91b2ob2o3bo$91bobo3bobo$86b2o6b5o$85bobo$85b2o5bo
3b3o$91bobobo3bo$91bobob3obo$88bo3bo5bob2o$87bobo7bo2bobo$88bo8bobo2bo
$91b3o2b2obobo$100bo22$164b2o$163b2o$165bo28$14b2o$15b2o$4b2o8bo$3bobo
$5bo3$12b2o$13b2o$12bo131bo35b2o$143b2o34b2o$143bobo35bo$14bo219bo$14b
2o168b2o47b2o$13bobo167b2o48bobo$185bo3$232bo$231b2o$231bobo2$177b2o$
177bobo$154b3o20bo$15b3o136bo$17bo137bo$16bo163b2o$180bobo$2o178bo$b2o
$o$14b2o$15b2o$14bo66$100b3o$100bo2bo$100bo$100bo$101bo!
However, we can synthesize them all in 32 gliders. Explicit syntheses are left to the reader.
Step back from that cliff, you idiots! Yes, you! Every last one of you! We wouldn't want to do something regrettable, do we now?

User avatar
dvgrn
Moderator
Posts: 8931
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Shinjuku: a database of glider syntheses

Post by dvgrn » July 28th, 2020, 8:49 pm

toroidalet wrote:
July 28th, 2020, 7:25 pm
The 239-glider synthesis doesn't work (one of the gliders hits the spider)...
Thanks for checking! Do you have a pattern handy for the failing 239-glider synth for the second spider? I take it that adding a few kickbacks and/or mirroring the recipe isn't likely to help?

User avatar
toroidalet
Posts: 1328
Joined: August 7th, 2016, 1:48 pm
Location: My computer
Contact:

Re: Shinjuku: a database of glider syntheses

Post by toroidalet » July 28th, 2020, 8:54 pm

Here:

Code: Select all

x = 98, y = 141, rule = B3/S23
61b2o$61b3o$60bob2o$60b3o$61bo9$35bo$33bo2bo2$32bo3b2o$33b5o$34bo$33b
6o$25bo7bo3bo$23bobo7bo3bo$24b2o11bo$11bo23bob2o$9bobo23bo2bo$10b2o22b
5o51bobo$34b2obo52b2o$25bo9b2o54bo$26b2o38bo$25b2o8b2o27b2o$34b2obo27b
2o20bobo$34b5o48b2o$35bo2bo49bo$35bob2o$37bo$33bo3bo$33bo3bo$33b6o$34b
o$24bo8b5o$25bo6bo3b2o$23b3o$33bo2bo17bo40bobo$35bo17bo41b2o$53b3o40bo
$23bo$21bobo$18bo3b2o$16bobo$17b2o$21bobo$22b2o$22bo$70bobo$70b2o9bobo
$71bo9b2o7bo$52bo7bo13b2o6bo5b2o$52bo3b2obobo12bobo12b2o$48bo3bo4bobo
2bo11bo$47bobo7bo2bobo$48bo3bo5bob2o$51bobob3obo$51bobobo3bo$45b2o5bo
3b3o$45bobo$46b2o6b5o$51bobo3bobo$51b2ob2o3bo$54bob3o$54bobo$53b2o$54b
obo$54bob3o$51b2ob2o3bo$51bobo3bobo$46b2o6b5o$45bobo$45b2o5bo3b3o$51bo
bobo3bo$51bobob3obo$48bo3bo5bob2o$47bobo7bo2bobo$48bo3bo4bobo2bo$52bo
3b2obobo$52bo7bo2$24b2o$23bobo$14bo10bo$14b2o$13bobo3$22b2o$21bobo$23b
o66b2o$53b3o34bobo$53bo36bo$54bo$23b3o68b2o$25bo68bobo$24bo69bo7$88bo$
87b2o$65b2o20bobo$25b2o37b2o$26b2o38bo$25bo65bo$90b2o$10b2o78bobo$9bob
o$11bo$24b2o$23bobo$25bo$b2o$obo$2bo13$61bo$60b3o$60bob2o$61b3o$61b2o!
It doesn't look like there's enough time for kickbacks, but maybe we could replace the troublesome gliders with another collision.
Step back from that cliff, you idiots! Yes, you! Every last one of you! We wouldn't want to do something regrettable, do we now?

User avatar
Ian07
Posts: 808
Joined: September 22nd, 2018, 8:48 am

Re: Shinjuku: a database of glider syntheses

Post by Ian07 » August 15th, 2020, 12:52 am

Finally got around to fixing up those two syntheses I posted here back in March; however it seems that one of them still refuses to work even though it only has one xWSS and is perfectly rewindable AFAIK:

Code: Select all

x = 0, y = 0, rule = B3/S23
32bo$33bo$31b3o37bo$27bo41b2o$28b2o40b2o$27b2o12$84bobo$84b2o$85bo9$2b
2o45b2o$2ob2o43bobo2b2o$4o44bobobobo$b2o44b2obo2bo$50bo$50b2o$10bo41bo
5b2o$8bobo41bo4bobo$9b2o42b2o3bo$5bo49bo$5b2o48b6o$4bobo53bo$57b3o$9bo
47bo$8b2o$8bobo12$86b2o$86bobo$86bo2$84bo$39bo43b2o$40b2o41bobo$39b2o
10$58b3o$58bo$59bo2$43b2o$44b2o$43bo3$57bo$57b2o5bo$56bobo4bo$63b3o2$
59b3o$59bo$60bo!
# Filtering with radius 5
# Engulfing yjggg0456zyk21y0361zypgy9265zyo223zzx4ocyzw46az48ezzzyog88cyzy31630g0gzyo1pl91344oyzws48013zyuok81v0s4yzxgocsozyy11yzy211zxgzw213zzydggyz0go8zyd211yv6a2x1
# processing src=xs38_caab8oz0652w1168ozy26520f871, dst=xp9_oo1v0c26zx344ox8a6zy311dd113...
# Invalid component
# Filtering with radius 8
# Filtering with radius 13
# Filtering with radius 21

Code: Select all

x = 236, y = 187, rule = B3/S23
28bo$29bo$27b3o3$128bobo$128b2o$129bo3$143bo$141b2o$142b2o2$29bo$30bo$
28b3o11$3bobo$4b2o$4bo2$2bo$obo$b2o12$78bobo$79b2o$79bo2$180bobo$180b
2o$181bo45$80bo$78b3o$77bo$77b6o$82bo$79bo3b2o$23b3o52bobo4bo$25bo52b
2o5bo$24bo61b2o$87bo$84bo2bob2o142b2o$83bobobobo142b4o$83b2o2bobo141b
2ob2o$87b2o143b2o15$146b2o$145b2o$147bo24$176b2o$176bobo$176bo15$3bo$
3b2o$2bobo12$158b2o$17b2o138b2o$18b2o139bo$17bo135b3o$153bo$154bo!
Anyone know what the issue is here?

User avatar
Ian07
Posts: 808
Joined: September 22nd, 2018, 8:48 am

Re: Shinjuku: a database of glider syntheses

Post by Ian07 » March 17th, 2021, 5:38 pm

The "prefer more steps" feature doesn't seem to actually work - I submitted some multi-step synths as a test which were equivalent to the single-steps that were already there, and nothing changed:

Code: Select all

x = 0, y = 0, rule = B3/S23
13bo$12bo$12b3o5$obo$b2o$bo$9bo$9bobo$4bobo2b2o$5b2o$5bo$17b2o$17b2o2$
17b2o$17b2o$5bo$5b2o$4bobo2b2o$9bobo$9bo$bo$b2o$obo5$12b3o$12bo$13bo!
# Filtering with radius 5
# Engulfing y8654z4oczy04ocw742zydrrzy0436ws48z436zy8ck4
# processing src=xs8_rr, dst=xs38_gs29e0e92sgz0349f0f943...
#     ...success.
# Filtering with radius 8
# Filtering with radius 13
# Filtering with radius 21
x = 0, y = 0, rule = B3/S23
42bo$42bobo$42b2o2$bo17bo$2bo14b2o$3o15b2o4$10bo$11b2o$10b2o2$20bo$20b
obo$20b2o3$19b2o$19b2o2$19b2o$19b2o3$20b2o$20bobo$20bo2$10b2o$11b2o$
10bo4$3o15b2o$2bo14b2o$bo17bo2$42b2o$42bobo$42bo!
# Filtering with radius 5
# Engulfing 0gydgyi742z223ya132zy6562y3gzyfgj21zyfddzyge24zy6532z2a6ya46azyzx712
# processing src=xs8_rr, dst=xs52_cimgehlmzxhaaaarz69d1ehld...
#     ...success.
# Filtering with radius 8
# Filtering with radius 13
# Filtering with radius 21
x = 0, y = 0, rule = B3/S23
4bobo$4b2o$5bo2$14bobo$14b2o$15bo2$2bo$bo$b3o3bo$6bo$6b3o$2o$2o2$2o$2o
$6b3o$6bo$b3o3bo$bo$2bo2$15bo$14b2o$14bobo!
# Filtering with radius 5
# Engulfing wgy7361zgj22wca8zddy0gggz06a2w12zy1gy4634zy0312
# processing src=xs8_rr, dst=xs26_3pmge2zg6q2sgz11...
#     ...success.
# Filtering with radius 8
# Filtering with radius 13
# Filtering with radius 21
(in hindsight, some of these could've been separated into even more steps, but again this is just a test)

https://catagolue.hatsya.com/object/xs3 ... f943/b3s23
https://catagolue.hatsya.com/object/xs5 ... ehld/b3s23
https://catagolue.hatsya.com/object/xs2 ... gz11/b3s23

User avatar
dvgrn
Moderator
Posts: 8931
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Shinjuku: a database of glider syntheses

Post by dvgrn » March 17th, 2021, 6:17 pm

Ian07 wrote:
March 17th, 2021, 5:38 pm
The "prefer more steps" feature doesn't seem to actually work ...
Yup, this was reported to have been changed quite a while ago -- in fact, I've finally found the reference, and you reported it back in mid-September -- but nobody has ever reported being able to get Catagolue to accept a same-cost more-steps recipe over a preexisting fewer-steps recipe.

Line 66 of diffupdate.py still says prefer_long = True, but maybe that's not a sufficient adjustment. There are certainly a lot of other dijkstra() calls elsewhere in the code that would end up using the default prefer_long = False. (I haven't tried to sort through the code nearly enough to have an educated opinion on this.)

User avatar
calcyman
Posts: 2591
Joined: June 1st, 2009, 4:32 pm

Re: Shinjuku: a database of glider syntheses

Post by calcyman » March 17th, 2021, 9:20 pm

The problem is with the Catagolue update process, rather than Shinjuku, and it's difficult to resolve without making deep changes to the way Catagolue stores glider syntheses.

Specifically, there's a census called b3s23/synthesis-costs which records the cost (in gliders) of every object Catagolue knows how to synthesise. The update process downloads this census, compares the glider counts against the ones in Shinjuku, and updates any syntheses in Catagolue which have strictly cheaper alternatives in Shinjuku.

Catagolue doesn't record the number of steps in the optimal glider syntheses it stores (it doesn't even know that intrinsically; it just stores a synthesis RLE), so the update process can only compare glider cost and not number of steps.

I think it might be possible to resolve this by including an auxiliary b3s23/synthesis-steps census to record the number of steps in each Catagolue synthesis; that way, the update process will have sufficient information to know when Shinjuku has an equal-cost synthesis with extra steps that should therefore be preferred by Catagolue.
What do you do with ill crystallographers? Take them to the mono-clinic!

User avatar
dvgrn
Moderator
Posts: 8931
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: Shinjuku: a database of glider syntheses

Post by dvgrn » March 17th, 2021, 9:29 pm

calcyman wrote:
March 17th, 2021, 9:20 pm
I think it might be possible to resolve this by including an auxiliary b3s23/synthesis-steps census to record the number of steps in each Catagolue synthesis; that way, the update process will have sufficient information to know when Shinjuku has an equal-cost synthesis with extra steps that should therefore be preferred by Catagolue.
Might there be a way to sort this out, that also makes it possible to find out somehow which syntheses make use of a given intermediate constellations? Some of the intermediates in the database can seem very mysterious otherwise.

User avatar
muzik
Posts: 4586
Joined: January 28th, 2016, 2:47 pm
Location: Scotland

Re: Shinjuku: a database of glider syntheses

Post by muzik » May 6th, 2021, 9:20 am

Is it just me or can infinite growth pattern syntheses just not be submitted? I've tried this about three times now, and while the RLE was considered completely valid nothing showed up after the given times, even though other submissions did.

User avatar
calcyman
Posts: 2591
Joined: June 1st, 2009, 4:32 pm

Re: Shinjuku: a database of glider syntheses

Post by calcyman » May 6th, 2021, 9:47 am

muzik wrote:
May 6th, 2021, 9:20 am
Is it just me or can infinite growth pattern syntheses just not be submitted? I've tried this about three times now, and while the RLE was considered completely valid nothing showed up after the given times, even though other submissions did.
Correct, they can't. If you want to include linear growth syntheses in Catagolue, you need to modify this file and submit a merge request:

https://gitlab.com/apgoucher/catagolue/ ... e.py#L6-21
What do you do with ill crystallographers? Take them to the mono-clinic!

User avatar
muzik
Posts: 4586
Joined: January 28th, 2016, 2:47 pm
Location: Scotland

Re: Shinjuku: a database of glider syntheses

Post by muzik » January 15th, 2022, 4:31 am

Any notability concerns aside, I'm amused by Catagolue's classification of this oscillator as "pseudo":
https://catagolue.hatsya.com/object/xp3 ... 62sg/b3s23

Post Reply