Page 18 of 24

Re: Catagolue Oddities

Posted: January 20th, 2019, 3:57 am
by Hunting
77topaz wrote:
Hunting wrote:Invalid pattern... What the heck?
http://catagolue.appspot.com/object/xp6 ... -ak3ais12a
The page is loading normally for me, I don't see any "invalid pattern" errors.
Oops, why?
EDIT: Oh, now I figured out. My network problem. That time the RLE is empty(Haven't been loaded) so invalid. Now okay.

Re: Catagolue Oddities

Posted: January 31st, 2019, 6:35 am
by 77topaz
A non-empty census that for some reason doesn't show a pie chart: http://catagolue.appspot.com/census/b36s245/D2_+2

I really don't understand this bug, because similar/related censuses don't show it.

Re: Catagolue Oddities

Posted: January 31st, 2019, 8:21 am
by calcyman
77topaz wrote:A non-empty census that for some reason doesn't show a pie chart: http://catagolue.appspot.com/census/b36s245/D2_+2

I really don't understand this bug, because similar/related censuses don't show it.
Pie charts were enabled on the 25th February 2015 (five days after Catagolue became operative). So there was an early 5-day window during which submitted hauls wouldn't contribute to pie charts, and censuses that haven't been updated since the 25th February 2015 therefore show no pie charts. The combination of very early investigation and subsequent complete inactivity is naturally very rare, which explains why you've only seen one census with this 'bug'.

Re: Catagolue Oddities

Posted: February 2nd, 2019, 3:31 am
by 77topaz
Currently, Catagolue appears to be partially down - the censuses are still viewable, and hauls can still be submitted, but no hauls are being committed for any census. (That suggests the cause may be something with the cron process?)

Re: Catagolue Oddities

Posted: February 2nd, 2019, 10:22 am
by calcyman
77topaz wrote:Currently, Catagolue appears to be partially down - the censuses are still viewable, and hauls can still be submitted, but no hauls are being committed for any census. (That suggests the cause may be something with the cron process?)
Thanks! Yes, I'd forgotten that Java's split() expects a regex rather than a substring, so it crashed on .split("+") and threw an exception.

Re: Catagolue Oddities

Posted: February 2nd, 2019, 8:10 pm
by mniemiec
calcyman wrote:Yes, I'd forgotten that Java's split() expects a regex rather than a substring, so it crashed on .split("+") and threw an exception.
From everything that I have read, it seems that split accepts either a regex or a string separator, and the behavior should depend on which you supply.

Re: Catagolue Oddities

Posted: February 2nd, 2019, 8:52 pm
by calcyman
mniemiec wrote:
calcyman wrote:Yes, I'd forgotten that Java's split() expects a regex rather than a substring, so it crashed on .split("+") and threw an exception.
From everything that I have read, it seems that split accepts either a regex or a string separator, and the behavior should depend on which you supply.
I mean that if you pass a string, it's treated as a regex: https://stackoverflow.com/a/24775872/5130486

Re: Catagolue Oddities

Posted: February 2nd, 2019, 9:38 pm
by mniemiec
calcyman wrote:I mean that if you pass a string, it's treated as a regex: https://stackoverflow.com/a/24775872/5130486
"this|that|other".split("|") produces "this","that","other" and "this.+.that.+.other".split("|") produces "this",".that",".other" in both Firefox and Chrome, which is exactly what one would expect if the split parameter is treated as a string and not a regular expression. I don't know what context that stackexchange article is using, but no specification I've seen for split says that a string parameter is converted to a regular expression, and given the above examples, it appears that it isn't. I've tried several other examples, and all confirm the string-only behavior.

Re: Catagolue Oddities

Posted: February 2nd, 2019, 10:11 pm
by calcyman
mniemiec wrote:
calcyman wrote:I mean that if you pass a string, it's treated as a regex: https://stackoverflow.com/a/24775872/5130486
"this|that|other".split("|") produces "this","that","other" and "this.+.that.+.other".split("|") produces "this",".that",".other" in both Firefox and Chrome
...in which language? Catagolue is written in Java, not JavaScript.

Re: Catagolue Oddities

Posted: February 2nd, 2019, 10:18 pm
by mniemiec
calcyman wrote:...in which language? Catagolue is written in Java, not JavaScript.
Oops! Sorry. Never mind (duh!)

Re: Catagolue Oddities

Posted: February 6th, 2019, 4:44 pm
by 77topaz
What's up with this pattern's Catagolue name ("063605590142055c0630")?

Re: Catagolue Oddities

Posted: February 7th, 2019, 5:16 am
by calcyman
77topaz wrote:What's up with this pattern's Catagolue name ("063605590142055c0630")?
It must have been imported from Andrzej's census. I've now renamed it to 'dead spark coil siamese eleven loop'.

Re: Catagolue Oddities

Posted: February 7th, 2019, 8:51 pm
by Hdjensofjfnen
calcyman wrote:
77topaz wrote:What's up with this pattern's Catagolue name ("063605590142055c0630")?
It must have been imported from Andrzej's census. I've now renamed it to 'dead spark coil siamese eleven loop'.
Fancy component in one of the soups:

Code: Select all

x = 13, y = 14, rule = B3/S23
6b3o$6bo$5b3o7$2o3b2o$obobobo$2bobo$obobobo4b2o$2o3b2o4b2o!
EDIT: So, hopefully, a 10G or less?

Code: Select all

x = 28, y = 35, rule = B3/S23
26bo$25bo$25b3o19$6b3o$6bo$5b3o7$2o3b2o$obobobo$2bobo$obobobo4b2o$2o3b
2o4b2o!
EDIT: 11G, annoyingly:

Code: Select all

x = 64, y = 81, rule = LifeHistory
60.A$59.A$59.3A37$17.A.A$17.2A$18.A7$17.3A$17.A$18.A$3A$2.A$.A13.3D$
15.D$14.3D7$9.2E3.2E7.A$9.E.E.E.E6.A$11.E.E8.3A$9.E.E.E.E4.2D$9.2E3.
2E4.2D3$25.3A$25.A$26.A7$62.2A$61.2A$63.A!
EDIT: :( Still 11G:

Code: Select all

x = 43, y = 46, rule = LifeHistory
3.A$4.A$2.3A8$.A$2.A38.A$3A37.A$18.A.A19.3A$18.2A$19.A7$18.3A$18.A$
19.A3$16.3D$16.D$15.3D7$10.2E3.2E7.A$10.E.E.E.E6.A$12.E.E8.3A$10.E.E.
E.E4.2D$10.2E3.2E4.2D3$26.3A$26.A$27.A!

Re: Catagolue Oddities

Posted: February 8th, 2019, 3:12 pm
by A for awesome
This 2x128 soup where a glider collision prolongs its lifespan by over 2^14 gens is marked as pathological:

Code: Select all

x = 128, y = 2, rule = B3/S23
obboobbbbbobboboooobbobboboboooobbboboboooboboobbbbbbbobobbbbobbbobboobobbbooobboooooobbboobbobooboboobboobooooboboobobbbbbobbob$
obooboobbboboobobbbboboobbobbobbbbobbbobobooobbooooobbbboooobbbboooooboobbbobooobboboboboobooobooobbbooobobbobbbobobbooooobboboo!

Re: Catagolue Oddities

Posted: February 20th, 2019, 1:46 pm
by Hdjensofjfnen
A bug in Catagolue allows still-life first-twenty-occurrences under 30 cells to be counted in discoveries if you go to the "attribute" page of the still life. Can someone fix this by having an automatic bot go to the "attribute" page of the SL and limit SL discoveries to the first five occurrences?

Re: Catagolue Oddities

Posted: February 20th, 2019, 3:08 pm
by calcyman
Hdjensofjfnen wrote:A bug in Catagolue allows still-life first-twenty-occurrences under 30 cells to be counted in discoveries if you go to the "attribute" page of the still life. Can someone fix this by having an automatic bot go to the "attribute" page of the SL and limit SL discoveries to the first five occurrences?
Why is this a bug? The idea is that if anyone calls /attribute on an object, that must be interesting to someone, and therefore worth including in users' discoveries.

Re: Catagolue Oddities

Posted: February 20th, 2019, 10:09 pm
by cordership3

Re: Catagolue Oddities

Posted: February 20th, 2019, 10:55 pm
by 77topaz
cordership3 wrote:'Very very very very boat': https://catagolue.appspot.com/object/xs ... 6521/b3s23
Odd: that's not the name of the LifeWiki page it leads, nor has it ever been the name of that LifeWiki page.

Re: Catagolue Oddities

Posted: February 21st, 2019, 7:09 am
by Ian07
77topaz wrote: Odd: that's not the name of the LifeWiki page it leads, nor has it ever been the name of that LifeWiki page.
It was, however, formerly the name used in the infobox which IIRC is what Catagolue actually looks at: http://conwaylife.com/w/index.php?title ... ldid=51091

Re: Catagolue Oddities

Posted: February 21st, 2019, 10:22 am
by calcyman
Ian07 wrote:
77topaz wrote: Odd: that's not the name of the LifeWiki page it leads, nor has it ever been the name of that LifeWiki page.
It was, however, formerly the name used in the infobox which IIRC is what Catagolue actually looks at: http://conwaylife.com/w/index.php?title ... ldid=51091
That's only part of the mystery. It's actually quite interesting to see how that hilarious name appeared on a Catagolue page. It all started when Apple Bottom abbreviated 'very very very very long' to 'very^4' on 1st July 2017, accidentally omitting the 'long' in the process. As AB did over 700 (!!!) maintenance edits on that day, it is understandable that one subtle error slipped through:

http://conwaylife.com/w/index.php?title ... ldid=34819

That, on its own, didn't look too incongruous -- the '^4' acts as something of a punctuation mark, splitting and therefore disguising the ungrammatical 'very boat'. There it lay dormant for nearly a year and a half, when AwesoMan3000 replaced 'very^4' with the equivalent 'very very very very' on 14th December 2018:

http://conwaylife.com/w/index.php?title ... ldid=51053

Again, this was part of a huge batch-editing bout, which may have even been partially automated: a find/replace to expand instances of 'very^4' to 'very very very very' sounds entirely reasonable. The problem, however, was that in this case, the original 'very^4 boat' was ungrammatical; this became especially apparent when changed to 'very very very very boat'.

The problem was finally resolved by a third serial batch-editor, Ian07, who corrected the name to agree with the page title whilst adding helpful statistics to many pattern pages. This removed the last trace of 'very very very very boat' from the current edition of the LifeWiki:

http://conwaylife.com/w/index.php?title ... ldid=51091

So the problem was completely solved, right? Wrong. The Python script used to scrape common names from LifeWiki and add them to the Catagolue source code was executed on 21st December, seven days after AwesoMan3000 changed the infobox to say 'very very very very boat' and a month before Ian07 corrected it:

https://gitlab.com/apgoucher/catagolue/ ... 7a6e02480f

This managed to evade detection -- unsurprisingly, given that the commit had a 2000-line diff -- until it was finally noticed yesterday.

Re: Catagolue Oddities

Posted: February 21st, 2019, 12:41 pm
by testitemqlstudop
LifeWiki wrote:Terribly long boat (or extra^3 boat) is a fairly large natural still life.
Congratulations, the next name for "very very very very boat" is "extra extra extra boat".

Re: Catagolue Oddities

Posted: February 21st, 2019, 2:20 pm
by dani
I vote to keep 'very very very very boat' as the name- it implies that it's more boat than anything previous to it.

Re: Catagolue Oddities

Posted: February 21st, 2019, 2:30 pm
by Moosey
danny wrote:I vote to keep 'very very very very boat' as the name- it implies that it's more boat than anything previous to it.
I guess I agree— it is more of a boat than any shorter boats.

Though perhaps then it would be boatyboat.

Re: Catagolue Oddities

Posted: February 21st, 2019, 4:25 pm
by Ian07
testitemqlstudop wrote:
LifeWiki wrote:Terribly long boat (or extra^3 boat) is a fairly large natural still life.
Congratulations, the next name for "very very very very boat" is "extra extra extra boat".
Wouldn't that just be extra^2?

extra = long^2
extra^2 = (long^2)^2 = long^(2*2) = long^4
Moosey wrote: I guess I agree— it is more of a boat than any shorter boats.

Though perhaps then it would be boatyboat.
Boaty McBoatFace?

Re: Catagolue Oddities

Posted: February 21st, 2019, 4:50 pm
by Moosey
Ian07 wrote:
testitemqlstudop wrote:
LifeWiki wrote:Terribly long boat (or extra^3 boat) is a fairly large natural still life.
Congratulations, the next name for "very very very very boat" is "extra extra extra boat".
Wouldn't that just be extra^2?

extra = long^2
extra^2 = (long^2)^2 = long^(2*2) = long^4
Vessel mathematics is super unintuitive, apparently:
In reality if something is extra^x it is very^(x+1)
Ian07 wrote:
Moosey wrote: I guess I agree— it is more of a boat than any shorter boats.

Though perhaps then it would be boatyboat.
Boaty McBoatFace?
Sure, why not? Then we could write a book about various parts of the history of CGoL and call it “the evolution of the Catagolue name for terribly long boat (and other miscellaneous oddities)”.