Pattern viewer for forum threads

[muzik]

Will the general-range outer-totalistic rules be able to support a neighbourhood size of 0, like WolframAlpha can for 1D rules?

----

Triangle neighbourhood on a hex grid (NT):

Code: Select all

x = 23, y = 10, rule = bshhistory
13.A$13.2A$5.A7.3A$5.2A6.4A$A4.3A5.5A$2A3.4A4.6A$AEA2.2AE2A3.3AE3A$4A
.6A2.8A$5.7A.9A$13.10A!
[[ VIEWONLY GRID THEME Blues ]]

Triangle neighbourhood on a triangular grid (ND):

Odd range values(?) (here R1, R3, R5?) would be as follows:

Code: Select all

x = 31, y = 8, rule = bslhistory
16.15A$5.9A3.13A$AEA3.3AE3A5.5AE5A$.A5.5A7.9A$8.3A9.7A$9.A11.5A$22.3A
$23.A!
[[ VIEWONLY GRID THEME Blues ]]

Even range values(?) (here R2, R4, R6?) would use the following:

Code: Select all

x = 44, y = 11, rule = bslhistory
$25.19A$10.13A3.17A$.7A3.11A5.15A$2.2AE2A5.4AE4A7.6AE6A$3.3A7.7A9.11A
$4.A9.5A11.9A$15.3A13.7A$16.A15.5A$33.3A$34.A!
[[ VIEWONLY GRID THEME Blues ]]

Looking back, it'd probably be better (if a bit weird) if the specified range would define what grid is used.

- Range 3n: use the hexagonal case
- Range 3n+1: use the triangular case where the central cell is flipped from the orientation of the whole triangle
- Range 3n+2: use the triangular case where the central cell and entire triangle have the same orientation

All would be NT.

[muzik]

I don't think LifeViewer is running this rule as expected, shouldn't it be throwing an error since I'm pretty sure PCA doesn't support B0 rather than apparently trying to run this as a Generations rule?

Code: Select all

x = 7, y = 3, rule = 2PCA4,15,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15
o3b3o$3o2bo$bo!

The forum software seems to be correctly rejecting it, but pasting it into lazyslug.com (build 618) shows the odd behaviour.

The 16 states does appear to be constant, so it's definitely recognising it as being a PCA rulestring but doing some arcane screwing up somewhere else such that it sets it to 16-state B3/S23.

Code: Select all

x = 7, y = 3, rule = 2PCA4,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0
o3b3o$3o2bo$bo!

EDIT: seems this problem isn't restricted to PCA:

Code: Select all

x = 9, y = 5, rule = W-1
$bo3b3o$b3o2bo$2bo!

Code: Select all

x = 9, y = 5, rule = W1
$bo3b3o$b3o2bo$2bo!

Code: Select all

x = 9, y = 5, rule = W256
$bo3b3o$b3o2bo$2bo!

Code: Select all

x = 9, y = 5, rule = W4294967300
$bo3b3o$b3o2bo$2bo!

Code: Select all

x = 9, y = 5, rule = B12345678/S239
$bo3b3o$b3o2bo$2bo!

Code: Select all

x = 9, y = 5, rule = it literally doesnt matter what i put here
$bo3b3o$b3o2bo$2bo!

Code: Select all

x = 9, y = 5, rule = lets try some commands that should throw an error
$bo3b3o$b3o2bo$2bo!
[[ HEXCELLS SQUARECELLS ]]

Minor fix-related suggestion: for invalid rules, instead of throwing a major error, display something closer to a script error instead and set the current rule to the "none" rule.

[muzik]

As can probably be told from the above report I've done a spot of thinking about B0/strobing rules and LifeViewer's support thereof. Currently there's quite a few rules that fall under the scary B0 umbrella that LifeViewer can't or won't simulate, for understandable reasons, but I'd prefer it if they could, and I've come up with some new mechanics which could allow for their safe implementation.

Currently for B0:
-common range-1 outer-totalistic, isotropic non-totalistic and non-isotropic rules with B0 are emulated via an alternating rule or black/white reversal, at all times
-general-range outer-totalistic rules are disallowed for unbounded grids, and are not emulated (i.e. they can strobe/truly B0) on bounded grids
-B0 common range-1 generations is disallowed
-B0 general-range generations is not disallowed and follows the same principles as the above outer-totalistic cases
-Margolus rules where something that is neither 0000 nor 1111 is birthed on 0000 are disallowed

According to my experimentation, it's possible to emulate general-range outer-totalistic rules (at least non-weighted ones) via an alternating rule in the exact same way you'd emulate range-1 rules, adjusting accordingly for the neighbourhood size. As such, I'd recommend implementing this for said rules on both bounded and unbounded grids.

Since strobing is unpleasant and potentially dangerous, all rules that can be emulated by alternating rules will be emulated, and those that cannot will throw up an error, regardless of whether or not the grid is bounded or not. If strobing is desired for some reason, a new script command called something along the lines of [[ NOEMULATE ]] would be implemented. If specified for unbounded grids, an error would be thrown up. If specified for bounded grids, a warning message regarding the strobe will be displayed upon opening the viewer (possibly alongside a mandatory delay for AUTOSTART, or possibly disallowing AUTOSTART entirely).

With NOEMULATE specified on a bounded grid:
-Common range-1 outer-totalistic, isotropic non-totalistic and non-isotropic rules with B0 will be played without emulating/converting them to another rule
-General-range outer-totalistic rules will also be played without emulation, like they currently do in LifeViewer and Golly
-Common range-1 Generations rules will be allowed to play
-General-range Generations will again be played without emulation, as is the case currently for LifeViewer and Golly
-Margolus rules which LifeViewer currently refuses to emulate will be playable without emulation, just like how MCell can
-PCA rules with similar B0 cases will be playable
-Odd Wolfram rules will be playable (if unusefully due to their current implementation, where all rows are ticked rather than only a single active one like MCell/Visions of Chaos do it, but it'll still be allowed, and alternating Wolfram rules are allowed anyway which have a similar problem so why not)
-Custom ruletables with a B0 element will be able to be directly simulated

[muzik]

On the topic of the previous post, after doing a bit of experimentation, I found out that Golly actually does emulate B0 now for general outer-totalistic (2 states). LifeViewer currently does not, so I assume this is on the backlog?

Code: Select all

x = 13, y = 13, rule = R1,C0,S0-3,B0,2-4,8
2bo$10b2o$4bo4bobo$8b3o$9bo$7bo3$10bo$bo$b2o9bo$4o$bo!

[rowett]

On the topic of the previous post, after doing a bit of experimentation, I found out that Golly actually does emulate B0 now for general outer-totalistic (2 states). LifeViewer currently does not, so I assume this is on the backlog?

Code: Select all

x = 13, y = 13, rule = R1,C0,S0-3,B0,2-4,8
2bo$10b2o$4bo4bobo$8b3o$9bo$7bo3$10bo$bo$b2o9bo$4o$bo!

True.

[goldenratio]

Copying RLE to clipboard doesn't seem to work in LifeViewer anymore. (it says it copied, but when I paste the clipboard it's still the same as before.) It seems to have been a problem for several days now (as A for awesome brought it up on Discord) but apparently nobody's reporting it until now.

[rowett]

Copying RLE to clipboard doesn't seem to work in LifeViewer anymore. (it says it copied, but when I paste the clipboard it's still the same as before.) It seems to have been a problem for several days now (as A for awesome brought it up on Discord) but apparently nobody's reporting it until now.

Which device and browser are you using?

[goldenratio]

Which device and browser are you using?

A Mac, and Google Chrome.

[muzik]

Identify seems to report inconsistent results for active cells upon identifying a predecessor of an oscillator and the oscillator itself, and it also seems to vary between normal/History and Super.

Code: Select all

x = 3, y = 2, rule = Life
.A$3A!

Code: Select all

x = 3, y = 2, rule = LifeHistory
.A$3A!

Code: Select all

x = 3, y = 2, rule = LifeSuper
.A$3A!

[yujh]

Of the first of these three patterns, lv said its volatility is 1.00

[rowett]

Identify seems to report inconsistent results for active cells upon identifying a predecessor of an oscillator and the oscillator itself, and it also seems to vary between normal/History and Super.

Fixed, thanks for reporting.

[muzik]

Identify seems to take longer to detect objects in Super than normal or History:

Code: Select all

x = 78, y = 90, rule = Life
5b2o$5b2o3$5bo$4b3o$3bo3bo$5bo18b2o$2bo5bo15bobo$2bo5bo10b2o4b3o12bo$3bo3bo7b
4o2bo4b3o8b4o4bo$4b3o8b3ob2o4b3o8b4o5bo$24bobo9bo2bo9b2o$24b2o10b4o9b2o$28bo
8b4o$27bo12bo$7bo19b3o3$7bo$5bobo$6b2o4$2o3b2o$14bo$bo3bo9bo$2b3o8b3o$2b3o3$
35bo$33b3o$3b2o27bo$3b2o27b2o7$20b2o$20bo$21b3o$23bo25$66b2o6b2o$65bo2bo4bo2b
o$64b6o2b6o$65bo2bo4bo2bo$66b2o6b2o8$59b3o$59bobo$59b3o$59b3o$59b3o$59b3o$59b
obo$59b3o!

Code: Select all

x = 78, y = 90, rule = LifeHistory
5b2o$5b2o3$5bo$4b3o$3bo3bo$5bo18b2o$2bo5bo15bobo$2bo5bo10b2o4b3o12bo$3bo3bo7b
4o2bo4b3o8b4o4bo$4b3o8b3ob2o4b3o8b4o5bo$24bobo9bo2bo9b2o$24b2o10b4o9b2o$28bo
8b4o$27bo12bo$7bo19b3o3$7bo$5bobo$6b2o4$2o3b2o$14bo$bo3bo9bo$2b3o8b3o$2b3o3$
35bo$33b3o$3b2o27bo$3b2o27b2o7$20b2o$20bo$21b3o$23bo25$66b2o6b2o$65bo2bo4bo2b
o$64b6o2b6o$65bo2bo4bo2bo$66b2o6b2o8$59b3o$59bobo$59b3o$59b3o$59b3o$59b3o$59b
obo$59b3o!

Code: Select all

x = 78, y = 90, rule = LifeSuper
5b2o$5b2o3$5bo$4b3o$3bo3bo$5bo18b2o$2bo5bo15bobo$2bo5bo10b2o4b3o12bo$3bo3bo7b
4o2bo4b3o8b4o4bo$4b3o8b3ob2o4b3o8b4o5bo$24bobo9bo2bo9b2o$24b2o10b4o9b2o$28bo
8b4o$27bo12bo$7bo19b3o3$7bo$5bobo$6b2o4$2o3b2o$14bo$bo3bo9bo$2b3o8b3o$2b3o3$
35bo$33b3o$3b2o27bo$3b2o27b2o7$20b2o$20bo$21b3o$23bo25$66b2o6b2o$65bo2bo4bo2b
o$64b6o2b6o$65bo2bo4bo2bo$66b2o6b2o8$59b3o$59bobo$59b3o$59b3o$59b3o$59b3o$59b
obo$59b3o!

For normal and History it takes 6258 generations, but for Super it takes 8080.

The traffic lights example also exhibits this to a degree of one extra generation.

[muzik]

Analysing the mysterious, unpredictable, no-blue-box slowdowns I seem to be getting now (generally after prolonged activity on one page), the frames seem to be reported as fluctuating around 40 to 50 fps, and a widely varying percentage as well.

On the prior topic of Identify: there's still a discrepancy between normal/History and Super for the traffic lights example, as it takes 15 rather than 13.

[rowett]

Analysing the mysterious, unpredictable, no-blue-box slowdowns I seem to be getting now (generally after prolonged activity on one page), the frames seem to be reported as fluctuating around 40 to 50 fps, and a widely varying percentage as well.

The throttling is most likely caused by Safari mobile. I'm unlikely to make any further progress since I don't have an iPad.

On the prior topic of Identify: there's still a discrepancy between normal/History and Super for the traffic lights example, as it takes 15 rather than 13.

Please post the pattern here.

[muzik]

On the prior topic of Identify: there's still a discrepancy between normal/History and Super for the traffic lights example, as it takes 15 rather than 13.

Please post the pattern here.

Code: Select all

x = 3, y = 2, rule = LifeHistory
.A$3A!

Code: Select all

x = 3, y = 2, rule = LifeSuper
.A$3A!

[martin.novy]

a bug?

when I go to viewtopic.php?p=128463#p128463

and click viewer on the NEXT comment,

the Lifeviewer
gave me:

invalid pattern
illegal character in pattern: <

but Lifeviewer at
viewtopic.php?p=129908#p129908
works ok

[rowett]

a bug?

I could not reproduce. Please give me a little more details on exactly which pattern you clicked on.

[martin.novy]

I could not reproduce. Please give me a little more details on exactly which pattern you clicked on.

Now I realized it was only because I use a browser extension called Diigo highlighting .

I apologize.

[hotdogPi]

Identify says this is a spaceship

Code: Select all

x = 8, y = 9, rule = LifeHistory
3.2F$2.4B$.6B$.6B$.B3A2B$FBABA2BF$.B3A2B$2.4B$2.B2F!

[rowett]

Identify says this is a spaceship

Code: Select all

x = 8, y = 9, rule = LifeHistory
3.2F$2.4B$.6B$.6B$.B3A2B$FBABA2BF$.B3A2B$2.4B$2.B2F!

It's because Identify is only looking a 2 states for [R]History patterns.

[cgoler2]

I request the feature that, when zoomed in to a specific zoom level, a grid is visible.

[bubblegum]

I request the feature that, when zoomed in to a specific zoom level, a grid is visible.

Press X, or the Grid key at the top-right, and make sure you're zoomed to at least 4.0.

[GUYTU6J]

A little question: why is the LifeViewer on the wiki page for adjustable glider loop lack of the Settings Menu buttom? Is it because of the POPUPWIDTH & POPUPHEIGHT commands?

[rowett]

A little question: why is the LifeViewer on the wiki page for adjustable glider loop lack of the Settings Menu buttom? Is it because of the POPUPWIDTH & POPUPHEIGHT commands?

It's probably to do with how big the browser window is. The settings button does not appear if the vertical height of the window is too small.

[cgoler2]

Is there any way to stop [[KILLGLIDERS]] from killing the glider that hits the LWSS and becomes a block in the twin hat modification soup?