Life tag testing

[rowett]

Golly [R]History rules...

I'd like them to be part of the same algo as the Super rules but the rule families have different cell colours and I can't figure out how to set the algo default colours once Golly has started (i.e. I need it dynamically on pattern load not statically when Golly starts or reads preferences). Otherwise when Golly starts it sets the Super state colours and then loading a History rule uses the wrong colours.

history.png (206.1 KiB) Viewed 211 times

[rowett]

Build 560 is now available on the lazyslug server

Enhancements in this build:

• new script command [[ EXCLUSIVEPLAY ]] will pause other Viewers when playback is started
• new script command [[ IGNOREEXCLUSIVE ]] will ignore pause requests from other Viewers

See the LifeViewer page on LifeWiki for an example of this in action. The top Viewer always plays. The other Viewers only play one at a time.

[rowett]

Build 561 is now available on the lazyslug server

Fixes in this build:

• fixed some bugs in [R]Super

See the LifeViewer page on LifeWiki for an example of this in action. The top Viewer always plays. The other Viewers only play one at a time.

[rowett]

Build 562 is now available on the lazyslug server

Fixes in this build:

• fixed B0 with native triangular rules

[dvgrn]

I've been having some trouble compiling Golly locally, so haven't been able to try out the Super/History implementation in Golly 4.0 yet.

I've been writing docs for [R]History and [R]Super, and figured I might as well quote the following paragraph here and make sure it was accurate:

With states 3 (white) and 5 (yellow), two additional colors are available to highlight ON cells that are special in some way.
State 5 was originally designed to carry information about cells that always remained ON from T=0 onward.
However, in practice this specialized functionality was seldom used, and state 5 became just an alternate color
to differentiate one type of highlighting from another. In LifeViewer, and now in Golly 4.0+, the yellow color
is permanently associated with any cell location where state 5 appears at T=0. If the cell in question turns OFF,
it changes to a red state-4 color. But if it turns back ON, the color will always be yellow.

This is certainly still true in LifeViewer, but I haven't tried it in the new Golly algo yet. The old LifeHistory rule table would turn the top and bottom yellow cells white after two ticks, but LifeViewer keeps them yellow:

Code: Select all

x = 9, y = 3, rule = LifeHistory
A3.C3.E$A3.C3.E$A3.C3.E!

That's true for the Golly implemention too, correct?

[rowett]

Build 563 is now available on the lazyslug server

Fixes in this build:

• fixed a bug in Hash neighbourhood

[rowett]

The old LifeHistory rule table would turn the top and bottom yellow cells white after two ticks, but LifeViewer keeps them yellow:

Code: Select all

x = 9, y = 3, rule = LifeHistory
A3.C3.E$A3.C3.E$A3.C3.E!

That's true for the Golly implemention too, correct?

The Golly [R]History implementation is the original LifeHistory rule table behaviour - so they go white after two ticks.

What would you prefer?

[dvgrn]

The Golly [R]History implementation is the original LifeHistory rule table behaviour - so they go white after two ticks.

What would you prefer?

Ah, and that's also true of the [R]Super implementation, both in LifeViewer and presumably in Golly: pink state-5 cells go white after two ticks.

Code: Select all

x = 9, y = 3, rule = LifeSuper
A3.C3.E$A3.C3.E$A3.C3.E!

I'm happy to keep the existing behavior, now that I'm sure what it is, and just document it.

However, it seems like there might be an argument in favor of making the LifeViewer [R]History behavior match its [R]Super behavior, since [R]Super is supposed to be an extension of [R]History. Originally you had very reasonable reasons for the minor difference in behavior, but now maybe it's not so necessary to maintain that discrepancy (?).

Minor Color Worries
The other [R]Super detail that I should probably look at one more time is the color scheme for LifeHistory-equivalent states. I just took muzik's palette suggestions without worrying about them too much, but now I've worked with them for a while... it doesn't seem as if there's a sufficiently good reason to change states 1-5 to colors that are unrecognizable as the old standard LifeHistory states.

Also state 21 is maybe a bit too difficult to tell apart from state 13. The other no-trail ON states are maybe about as good as they're going to get, but that one is kind of tough. When I'm describing a LifeSuper pattern it's difficult even to mentally assign a color name to state 21:

"In the last row, the orange switch engines alternate with switch engines that are, um, very pale slightly greenish yellow"

If we do tweak the colors a bit, I can't currently tell state 7 from state 3 by looking at it. And maybe the "orange" state 17 are also a bit too close to the state-15 color. But possibly I can get away with thinking of state 15 as "red-orange".

Do you think there's any hope of reclaiming something close to LifeHistory colors for the first five states? Maybe just pastelify the LIfeHistory colors a little bit? I guess that would cause trouble with state 19, so maybe that's a can of worms not worth opening.

[rowett]

However, it seems like there might be an argument in favor of making the LifeViewer [R]History behavior match its [R]Super behavior, since [R]Super is supposed to be an extension of [R]History. Originally you had very reasonable reasons for the minor difference in behavior, but now maybe it's not so necessary to maintain that discrepancy (?).

Agreed, I'll take a look at LifeViewer [R]History and see how easy it is to modify.

Do you think there's any hope of reclaiming something close to LifeHistory colors for the first five states? Maybe just pastelify the LIfeHistory colors a little bit? I guess that would cause trouble with state 19, so maybe that's a can of worms not worth opening.

It's trivial for me to change the default [R]Super colours in both LifeViewer and Golly so just let me know what you'd like. I'm not a fan of the [R]Super first five state colours I think they make the patterns look odd. But feel free to propose any colours for any of the states.

[dvgrn]

It's trivial for me to change the default [R]Super colours in both LifeViewer and Golly so just let me know what you'd like. I'm not a fan of the [R]Super first five state colours I think they make the patterns look odd. But feel free to propose any colours for any of the states.

Okay, this will go on my to-do list after a first draft of the Golly Super.html docs. Might not be for a few days -- the early part of this week is shaping up to be juggle-six-things busy as it is.

These aesthetic decisions about colors are fairly painful, and tend to get one into "can't please everyone" problems like in the Aesop's fable of the two men and the donkey. So I'll probably just make something up as quickly as I can, cross my fingers and hope. (And if any modifications of my modifications occur to you, I'll be very happy to take suggestions.)

[rowett]

Build 564 is now available on the lazyslug server

Fixes in this build:

• fixed a bug with initial AutoFit on hex patterns

[rowett]

Build 565 is now available on the lazyslug server

Fixes in this build:

• ensure paste can't create invalid states

[dvgrn]

However, it seems like there might be an argument in favor of making the LifeViewer [R]History behavior match its [R]Super behavior, since [R]Super is supposed to be an extension of [R]History. Originally you had very reasonable reasons for the minor difference in behavior, but now maybe it's not so necessary to maintain that discrepancy (?).

Agreed, I'll take a look at LifeViewer [R]History and see how easy it is to modify.

Any conclusion about this yet?

It's trivial for me to change the default [R]Super colours in both LifeViewer and Golly so just let me know what you'd like.

Okay, here's an attempt at an alternate LifeSuper colorization:

Code: Select all

@COLORS
1    0  255    0  # standard ON (green)
2    0    0  255  # history state OFF (blue)
3  255  216  255  # marked ON with trail (pink)
4  255    0    0  # marked OFF with trail (red)
5  255  255    0  # marked ON alternate (yellow)
6   96   96   96  # boundary cell (dark gray)
7  255    0  255  # marked ON no-trail filter (magenta)
8  128    0  128  # marked OFF no-trail filter (purple)
9    0  191  255  # ON with trail alt#1 (deep sky blue)
10   0   64  128  # OFF trail alt#1 (dark cerulean)
11  64  224  208  # ON with trail (turquoise)
12   0  128   64  # OFF trail alt#2 (bluish green)
13 255  255  255  # ON no-trail #1 (white)
14 255   99   71  # OFF disappearing label (tomato)
15 250  128  114  # ON no-trail #2 (salmon)
16 219  112  147  # OFF vanishing label (pale violet red)
17 255  165    0  # ON no-trail #3 (orange)
18 245  222  179  # OFF cycling label #1 (wheat)
19   0  255  255  # ON no-trail #4 (cyan)
20 150   50  204  # OFF cycling label #2 (dark orchid)
21 192  255  128  # ON no-trail #5 (yellow green) 
22 255  182  193  # OFF cycling label #3 (light pink)
23   0  255  127  # ON no-trail #6 (brighter green)
24   1    1    1  # hidden label (stealthy near-black)
25 255    0  127  # ON no-trail #7 (red magenta)

I think this keeps the flavor of muzik's colors, while restoring some of the functionality that [R]Super is going to need -- some states you can just think of as "state 15", "state 17", etc., but some are really key functional states, and you need to be able to think about a glider "going through a red boundary and then a purple boundary", not just "going through a greenish boundary that looks similar to State 12 but has a totally different function".

So... the no-trail barrier is now a definite purple/magenta color, not some confusing white/green pairing. Gliders with trails that go through a purple barrier are currently turning white. Originally I had them turning pink, but even I could tell that that was a painful color choice for such a common cell type. But then the marked ON color in LifeHistory was too close to white, o I changed that color color to pink instead -- a color that's close to the red of an OFF marked cell.

LifeHistory colors 1-6 are otherwise back to their old standards, for better or for worse -- except that the blue of State 2 is now a much brighter blue in LifeSuper. It seemed like a good idea to be able to tell Super apart from History, generally, pretty much at a glance -- and also it seems important for the "trail" state to not fade into the background so much. Now that it's possible to choose either trail or no-trail, presumably when you choose to show a trail, you'll probably want to see it clearly.

So white is is the default lowest-common-denominator no-trail state, and it's not painfully far away from the next few no-trail states, but you can easily identify it in a crowded pattern. Yellow and cyan had to get swapped around to avoid confusion with standard LifeHistory colors, but I don't think that came out too bad.

Stealth State 24
The big difference between Golly and LifeViewer is how well state 24 shows up. That was a deliberate difference, made possible by the dark gray default background state in Golly vs. plain black in LifeViewer. The "stealth labels" are supposed to be a bit easier to see and edit in Golly, whereas they're more likely to see actual use in LifeViewer animations, where it's good if they don't show up much until they're activated.

I could see maybe making them a little grayer, so that it's just possible to pick them out if you're looking for them -- but then again this might end up being unavoidably different on different monitors, so I'm pretty happy to just leave them as they are.

Nefarious Plans
I'll think this over for a day or so, try things out, see if you have any better ideas, and then bite the bullet and make the change to the LifeSuper rule on LifeWiki -- and open the floodgates for counter-arguments on the LifeSuper thread.

Nefarious Postscript:
Of course changing the colors in the LifeSuper rule table won't change the colors in LifeViewer any more. Hm. Whenever it seems like a good moment to release an updated LifeViewer build, I'll be ready to post something on the LifeSuper thread pointing out and attempting to justify the color changes. Probably it makes sense to try things out in LifeViewer for a little while, see if any more obvious improvements come in, before moving the colors into Golly.

[rowett]

Agreed, I'll take a look at LifeViewer [R]History and see how easy it is to modify.

Any conclusion about this yet?

Not yet I've been distracted with all things Golly. Will get to it soon.

Okay, here's an attempt at an alternate LifeSuper colorization:

I've implemented these in LifeViewer here.

I certainly prefer them to the original set.
Perhaps the state 2 is a little bright?

[dvgrn]

I've implemented these in LifeViewer here.

I certainly prefer them to the original set.
Perhaps the state 2 is a little bright?

I agree. I wanted it visibly different from LifeHistory blue, but that was back when that was the _only_ difference from LifeHistory. But now we also have the white -> pink shift, though that's surprisingly subtle in the context of the average pattern. The blue trail color should also be about the same brightness as the other two trail colors, and at (0 0 255) it's definitely not.

A little experimentation with switching back and forth between LifeHistory and LifeSuper says that (0 0 144) is not bright enough to be visibly different, and (0 0 192) is still too bright. So how about (0 0 160)?

[rowett]

So how about (0 0 160)?

Yes looks better - updated here.

[dvgrn]

Yes looks better - updated here.

Looks good to me. That might be about the best color set I can figure out without outside help.

My remaining worries are

1) State 1 is now too close to State 23. This doesn't matter terribly much, since one is a trailed state and the other is no-trail, so they're easy to distinguish by behavior. It's not ideal, but if I move state 23 in the obvious direction it will get too close to state 21. State 21 has an easy mental label (yellow-green) where state 23 doesn't so much (SlightlyMoreBluishGreen?).

However, I'm having a hard time finding any remaining empty spots in the color wheel among the no-trail states, with a name that a non-interior-decorator type person can use as a mental label. It seems okay to use a mental label of "green no-trail", and to have green represented as a no-trail state, along with yellow-green which is distinguishable from it but doesn't clash horribly.

Things like this seem like they'll be one of the more common uses of LifeSuper, and I think they still look okay with the color substitutions so far. The most similar colors are definitely the green and the yellow-green, and that seems more or less tolerable:

Code: Select all

x = 268, y = 267, rule = LifeSuper
4.W$5.2W248.W$4.2W247.2W$254.2W10.W$265.W$265.3W2$245.W$245.W.W$245.
2W9$250.S$17.S230.2S$18.2S229.2S$6.S10.2S$7.S$5.3S2$27.S$25.S.S$26.2S
9$22.O$23.2O212.O$22.2O211.2O$236.2O10.O$247.O$247.3O2$227.O$227.O.O$
227.2O9$232.U$35.U194.2U$36.2U193.2U$24.U10.2U$25.U$23.3U2$45.U$43.U.
U$44.2U9$40.pA$41.2pA176.pA$40.2pA175.2pA$218.2pA10.pA$229.pA$229.3pA
2$209.pA$209.pA.pA$209.2pA9$214.W$53.W158.2W$54.2W157.2W$42.W10.2W$
43.W$41.3W2$63.W$61.W.W$62.2W9$58.S$59.2S140.S$58.2S139.2S$200.2S10.S
$211.S$211.3S2$191.S$191.S.S$191.2S9$196.O$71.O122.2O$72.2O121.2O$60.
O10.2O$61.O$59.3O2$81.O$79.O.O$80.2O9$76.U$77.2U104.U$76.2U103.2U$
182.2U10.U$193.U$193.3U2$173.U$173.U.U$173.2U9$178.pA$89.pA86.2pA$90.
2pA85.2pA$78.pA10.2pA$79.pA$77.3pA2$99.pA$97.pA.pA$98.2pA9$94.W$95.2W
68.W$94.2W67.2W$164.2W10.W$175.W$175.3W2$155.W$155.W.W$155.2W9$160.S$
107.S50.2S$108.2S49.2S$96.S10.2S$97.S$95.3S2$117.S$115.S.S$116.2S9$
112.O$113.2O32.O$112.2O31.2O$146.2O10.O$157.O$157.3O2$137.O$137.O.O$
137.2O9$142.U$125.U14.2U$126.2U13.2U$114.U10.2U$115.U$113.3U2$135.U$
133.U.U$134.2U3$130.pA$3W15.3O15.3pA15.3S15.3U15.3W15.3O18.pA.pA12.3U
15.3S15.3pA15.3O15.3W15.3U15.3S$2.W17.O17.pA17.S17.U17.W17.O19.2pA12.
U17.S17.pA17.O17.W17.U17.S$.W17.O17.pA17.S17.U17.W17.O22.2pA11.U17.S
17.pA17.O17.W17.U17.S$9.3S15.3U15.3W15.3O15.3pA15.3S15.3U12.pA.pA18.
3O15.3W15.3U15.3S15.3pA15.3O15.3W$11.S17.U17.W17.O17.pA17.S17.U14.pA
2.2pA14.O17.W17.U17.S17.pA17.O17.W$10.S17.U17.W17.O17.pA17.S17.U14.UpA
.pA.pA15.O17.W17.U17.S17.pA17.O17.W$136.pA$133.2U.pA.pA$134.U2.2pA$
132.W.W$132.2W$130.2W$129.W.W$125.2W2.W$125.W.W.WU$127.W$125.W.W.2U$
125.2W2.U$129.S.S$130.2S!
#C [[ THUMBNAIL THUMBSIZE 2 AUTOSTART Y 104 Z 8 TRACKLOOP 36 0 -1/4 ]]

2) Similarly, "pink no-trail" is likely to be my inaccurate mental label for State 25, even though it's really "red magenta". For that matter, magenta state 7 will probably end up as "pink" in my mental models, or "pink boundary" -- three-syllable color names like "magenta" really just don't stick in my head.

Hot pink might work better than magenta for state 7, anyway. What do you think of (255 105 180)? This is a transient state -- it will rarely be there in a T=0 pattern -- but to me that shade gives a good sense of overlap between the boundary-OFF purple and the white state-13 which is the usual color of a spaceship coming out past that boundary.

Code: Select all

x = 109, y = 27, rule = LifeSuper
8D23.5D21.3D6.3D18.3D4.6D8.H$5.3D22.3D.3D18.5D4.5D16.5D4.2D2.3D7.H$5.
2D22.3D3.3D17.5D4.5D16.5D8.3D7.H$4.3D22.3D3.3D19.3D6.3D18.3D8.3D7.H$
4.2D23.3D3.3D19.3D6.3D18.3D7.4D7.H$3.3D23.4D.4D19.3D6.3D18.3D4.6D8.H$
3.2D25.8D19.3D6.3D18.3D8.3D7.H$2.3D30.3D19.3D6.3D18.3D8.4D6.H$2.3D29.
3D20.3D6.3D18.3D8.4D6.H$2.2D29.4D20.3D6.3D18.3D3.3D2.3D7.H$.3D25.6D
20.8D.8D13.8D.6D8.H$108.H$108.H$108.H$108.H$108.H$108.H$108.H$.5G25.
5I25.5K25.5M12.H$G4.G24.I4.I24.K4.K24.M4.M12.H$5.G29.I29.K29.M12.H$G
3.G25.I3.I25.K3.K25.M3.M13.H$2.G29.I29.K29.M15.H$108.H$108.H$108.H$
108.H!
#C [[ THUMBNAIL THUMBSIZE 3 ]]

3) The "dark orchid" label color, state 20, is really just another purple, and also it doesn't really fit with the rest of the palette. The purple no-trail boundary color that I've invented is now a major competing use for the mental label "purple". Is there a better choice for that member of the three "cycling" colors?

This doesn't seem to matter too much, since I would hope that those label colors wouldn't get used very often anyway. Like BLINK tags in old HTML, they were probably just a bad idea, brought on by the fact that I had lots of extra OFF label states and a shortage of ideas for functionality for them.

[rowett]

Hot pink might work better than magenta for state 7, anyway. What do you think of (255 105 180)?

I definitely prefer it.

3) The "dark orchid" label color, state 20, is really just another purple, and also it doesn't really fit with the rest of the palette.

How about 96, 240, 160? It's on the ramp between 19 and 21. Might be too similar to 21 but fits better in the palette.

You can try both here.

Also I looked at the LifeViewer [R]History algo. Basically it requires that the marked states are constant (and modified at display time based on the underlying cell being dead or alive) and do not change over generations which is why the 5->4->5 final transition goes back to 5 rather than to 3.

I can switch it to use a variant of the [R]Super algo:
pros:
1) it will consistent with Golly
2) it will actually simplify a fair amount of code (both algo and rendering)
cons:
1) it will be up to 10x slower in the extreme cases simulating [R]History patterns
(note: [R]Super is also up to 10x slower in the extreme cases than the current [R]History implementation)

Thoughts?

[dvgrn]

How about 96, 240, 160? It's on the ramp between 19 and 21. Might be too similar to 21 but fits better in the palette.

It seems somewhat workable, but might be moving the color scheme a bit too far towards pink. When I look at the current lazyslug LifeSuper palette, I see some similarity between 3, 7, 16, 22, 25 -- particularly between 16 and 22.

Those are both OFF colors, so I'd rather have them be easily distinguishable if possible.

The only place on the color wheel that doesn't seem overused is the white-to-black continuum -- we just have white, dark gray, and black. How about if state #20 is just something like silver, (192 192 192)?

That looks okay to me, not too garish, not too close to State 6 which mostly won't get used in normal patterns anyway. The three cycling colors don't look too bad together, maybe a little better than with the purple even. The "crazy labels" aren't quite so flickery.

And the nice thing is that if somebody really really wants purple labels, they can perfectly well use state 8 -- you just can't fire spaceships through the labels to destroy them. Instead the spaceships will get through but their color(s) will get bleached out.

Code: Select all

x = 396, y = 223, rule = LifeSuper
2.3D26.5D25.6D23.8D23.5D21.3D6.3D18.3D4.6D17.3D4.7D16.3D3.9D15.3D5.5D
16.5D6.3D16.5D4.5D16.5D3.7D$5D25.2D2.3D24.2D32.3D22.3D.3D18.5D4.5D16.
5D4.2D2.3D14.5D4.3D18.5D9.3D13.5D4.3D.3D14.2D2.3D3.5D15.2D2.3D2.2D3.
2D14.2D2.3D2.3D$5D29.3D24.2D32.2D22.3D3.3D17.5D4.5D16.5D8.3D14.5D4.3D
18.5D8.3D14.5D3.3D3.3D18.2D3.5D20.2D7.2D19.2D2.3D$2.3D29.3D24.2D31.3D
22.3D3.3D19.3D6.3D18.3D8.3D16.3D4.3D20.3D8.3D16.3D3.3D3.3D18.2D5.3D
20.2D7.2D19.2D2.3D$2.3D29.2D25.6D27.2D23.3D3.3D19.3D6.3D18.3D7.4D16.
3D4.6D17.3D7.3D17.3D3.3D3.3D18.2D5.3D19.3D6.3D18.3D2.6D$2.3D26.4D30.
3D25.3D23.4D.4D19.3D6.3D18.3D4.6D17.3D8.3D16.3D7.3D17.3D3.4D.4D17.3D
5.3D19.3D3.5D19.2D8.2D$2.3D29.3D29.2D25.2D25.8D19.3D6.3D18.3D8.3D16.
3D9.3D15.3D6.4D17.3D4.8D16.3D6.3D18.3D8.3D16.3D8.3D$2.3D30.2D29.2D24.
3D30.3D19.3D6.3D18.3D8.4D15.3D9.3D15.3D6.3D18.3D9.3D15.3D7.3D17.3D9.
3D15.3D9.3D$2.3D30.2D28.3D24.3D29.3D20.3D6.3D18.3D8.4D15.3D8.4D15.3D
6.3D18.3D8.3D15.3D8.3D16.3D10.3D14.3D10.3D$2.3D25.2D2.3D23.2D3.2D25.
2D29.4D20.3D6.3D18.3D3.3D2.3D16.3D3.3D2.3D16.3D5.3D19.3D7.4D14.3D9.3D
15.3D6.2D3.3D13.3D6.2D2.3D$8D22.6D25.5D25.3D25.6D20.8D.8D13.8D.6D15.
15D15.8D2.3D17.14D16.8D2.7D13.8D2.5D15.8D.6D2$386.2A$386.A.A.2A.A$
388.A.A.2A$388.2A3$.5A25.5C25.5E25.5G25.5I25.5K25.5M25.5O25.5Q25.5S
25.5U25.5W25.5pA25.2A$A4.A24.C4.C24.E4.E24.G4.G24.I4.I24.K4.K24.M4.M
24.O4.O24.Q4.Q24.S4.S24.U4.U24.W4.W24.pA4.pA25.2A$5.A29.C29.E29.G29.I
29.K29.M29.O29.Q29.S29.U29.W29.pA$A3.A25.C3.C25.E3.E25.G3.G25.I3.I25.
K3.K25.M3.M25.O3.O25.Q3.Q25.S3.S25.U3.U25.W3.W25.pA3.pA$2.A29.C29.E
29.G29.I29.K29.M29.O29.Q29.S29.U29.W29.pA10$166.3V9.2V31.3V$166.3V9.
2V31.3V$166.3V9.2V31.3V$166.3V9.2V31.3V$162.7V2.6V.2V2.5V.4V3.2V2.5V
3.7V$161.3V.4V.3V.6V.2V2.3V.3V3.2V.2V2.3V.3V.4V$160.3V3.6V3.5V6.2V2.
2V2.5V4.5V3.3V$160.3V3.14V2.6V2.3V.2V.11V3.3V95.7N58.4N$160.3V3.6V6.
2V.2V3.2V3.2V.2V.2V6.3V3.3V95.7N58.4N$160.3V3.6V6.2V.2V3.2V3.4V2.3V5.
3V3.3V95.3N$161.3V.4V.3V2.5V.2V2.3V3.4V3.2V3.2V.3V.4V95.3N$161.8V2.6V
.3V.7V3.3V4.5V2.8V91.7N.3N.5N3.5N3.7N2.7N3.6N3.5N3.9N.7N2.8N$192.2V
110.3N.4N.6N.3N.2N2.3N2.3N.4N.3N.4N.3N.4N.2N2.3N2.3N3.3N.3N.3N.3N2.3N
$191.3V109.3N3.3N.6N9.3N2.3N2.3N.3N2.6N3.3N5.3N2.3N3.3N.2N3.6N2.3N$
191.3V109.3N3.3N.3N.4N3.7N2.2N3.3N.2N3.12N.7N2.2N4.3N.2N3.6N2.3N$191.
V111.3N3.3N.3N3.8N.3N2.2N3.3N.2N3.6N6.4N.3N2.2N4.3N.2N3.10N$168.5V.5V
.5V2.10V3.2V101.3N3.3N.3N4.6N2.3N2.3N2.3N.3N2.6N6.3N2.3N2.2N4.3N.2N3.
6N$167.3V2.V.3V2.2V2.3V5.2V.3V3.2V102.3N.4N.5N2.6N2.3N2.3N.3N.4N.3N2.
3N2.6N2.3N2.2N4.3N.2N3.11N$167.2V5.2V8.2V4.3V2.2V2.3V102.8N.3N.5N2.
16N2.7N3.6N2.11N4.3N.2N3.6N2.4N$166.3V5.2V4.6V3.3V3.3V.2V132.2N7.2N
43.3N3.3N$166.3V5.2V3.2V3.2V3.2V5.2V.2V132.2N7.2N43.3N2.3N$167.2V5.2V
3.2V3.2V2.3V5.4V133.2N7.2N44.6N$167.3V2.V.2V3.2V2.3V2.2V6.4V$168.8V3.
15V3.3V$197.2V$196.3V$196.3V126.3N10.2N16.3N$166.2V9.2V15.2V129.3N10.
2N16.3N$166.2V9.2V15.2V129.3N10.2N16.3N$166.2V9.2V15.2V129.3N10.2N16.
3N$166.2V9.2V15.2V129.3N2.5N3.7N3.6N2.3N.5N$166.2V2.5V2.7V3.5V2.2V.5V
123.3N.2N2.3N2.3N.4N.3N.4N.6N.3N$166.2V.2V2.3V.3V2.3V.2V2.3V.2V.2V2.V
123.3N5.3N2.3N2.6N3.3N.6N$166.2V6.2V.2V4.4V4.2V.2V.2V126.3N.7N2.2N3.
12N.3N.4N$166.2V2.6V.2V4.10V.2V.4V124.7N.3N2.2N3.6N7.3N3.4N$166.2V.2V
3.2V.2V4.4V7.2V3.3V123.6N2.3N2.3N2.6N7.3N4.3N$166.2V.2V3.2V.2V4.5V6.
2V4.3V122.6N2.3N2.3N.3N2.3N2.3N.5N2.3N$166.2V.2V2.3V.3V2.3V.2V3.2V.4V
2.3V122.3N.16N3.6N2.3N.5N$166.3V.14V3.5V.3V.5V$2.164V$.V$2.164V$166.V
$2.164V$.V$2.164V154.20D$166.V$2.164V$.V$2.164V$166.V$2.164V98.3P61.
2P7.3P$.V262.3P61.2P7.3P$2.164V98.3P17.3P26.3P12.2P7.3P$166.V97.3P17.
3P26.3P12.2P7.3P$2.83R81V94.7P2.6P2.5P.6P.8P3.3P2.5P.6P2.5P3.7P2.3P2.
6P$.R164.V92.3P.4P.3P.7P.3P.3P3.3P2.3P3.3P.3P.3P.3P3.2P2.3P2.3P.4P.3P
.3P.4P$2.164T92.3P3.6P3.6P5.3P3.3P2.3P3.6P6.3P7.3P2.3P2.3P.6P3.3P$
258.3P3.12P.4P3.3P3.2P3.3P3.6P6.3P3.7P2.2P3.3P.12P$258.3P3.6P9.4P.3P
3.2P3.3P3.6P6.3P2.4P.3P2.2P3.3P.6P$258.3P3.6P10.3P.3P3.2P4.2P3.6P6.3P
2.3P2.3P2.3P2.3P.6P$259.3P.4P.3P2.5P2.3P.3P3.2P4.3P.4P.3P.3P.3P2.3P2.
3P2.3P.3P.4P.3P2.3P$259.8P2.6P2.5P3.7P4.8P2.5P3.4P.16P.4P2.6P6$280.3P
10.2P16.3P$280.3P10.2P16.3P$280.3P10.2P16.3P$280.3P10.2P16.3P$280.3P
2.5P3.7P3.6P2.3P.5P$280.3P.2P2.3P2.3P.4P.3P.4P.6P.3P$280.3P5.3P2.3P2.
6P3.3P.6P$280.3P.7P2.2P3.12P.3P.4P$280.7P.3P2.2P3.6P7.3P3.4P$280.6P2.
3P2.3P2.6P7.3P4.3P$280.6P2.3P2.3P.3P2.3P2.3P.5P2.3P$280.3P.16P3.6P.4P
.5P$170.5V.5V.5V2.10V3.2V$169.3V2.V.3V2.2V2.3V5.2V.3V3.2V$169.2V5.2V
8.2V4.3V2.2V2.3V$168.3V5.2V4.6V3.3V3.3V.2V$168.3V5.2V3.2V3.2V3.2V5.2V
.2V$169.2V5.2V3.2V3.2V2.3V5.4V78.20H$169.3V2.V.2V3.2V2.3V2.2V6.4V$
170.8V3.15V3.3V$199.2V$198.3V$198.3V4.2V$168.2V9.2V15.2V7.2V$168.2V9.
2V15.2V7.2V$168.2V9.2V15.2V7.2V$168.2V9.2V15.2V7.2V46.3R11.4R$168.2V
2.5V2.7V3.5V2.2V.5V.2V46.3R11.4R$168.2V.2V2.3V.3V2.3V.2V2.3V.2V.2V2.V
.2V46.3R$168.2V6.2V.2V4.4V4.2V.2V.2V4.2V46.3R$168.2V2.6V.2V4.10V.2V.
4V2.2V37.7R2.3R3.5R2.5R.6R$168.2V.2V3.2V.2V4.4V7.2V3.3V.2V37.3R.4R.3R
2.2R2.3R3.3R.3R2.2R$168.2V.2V3.2V.2V4.5V6.2V4.3V39.3R2.3R.3R6.3R3.3R.
3R3.2R$168.2V.2V2.3V.3V2.3V.2V3.2V.4V2.3V39.2R3.3R.3R2.7R3.3R.2R4.2R$
168.3V.14V3.5V.3V.5V.RT37.2R3.3R.3R.4R.3R3.3R.2R4.2R$205.VT37.3R2.3R.
3R.3R2.3R3.3R.2R4.2R$244.3R.3R2.3R.3R2.3R3.3R.2R4.2R$244.7R2.3R2.9R.
3R.2R4.2R$244.2R$244.2R$244.2R5$240.3R12.2R17.3R$240.3R12.2R17.3R$
240.3R12.2R17.3R$240.3R12.2R17.3R$240.3R3.5R4.7R4.6R2.3R2.5R$240.3R2.
2R2.3R3.3R.4R2.3R.4R.3R.3R.3R$240.3R6.3R3.3R2.3R.3R3.3R.3R.3R$240.3R
2.7R3.2R3.3R.9R.3R2.4R$240.3R.4R.3R3.2R3.3R.3R7.3R4.4R$240.3R.3R2.3R
3.3R2.3R.3R7.3R5.3R$210.3X8.2X17.3R.3R2.3R3.3R.3R3.3R2.3R.3R.2R2.3R$
210.3X8.2X17.3R2.9R.7R4.6R2.3R2.5R$186.3X21.3X2.2X4.2X$186.3X21.3X2.
2X4.2X$179.5X.6X2.5X4.5X3.9X2.6X$179.2X2.X2.3X3.3X2.2X2.2X3.2X2.3X2.
2X4.3X.3X$179.2X5.3X3.2X3.3X6.2X2.3X2.2X4.2X3.2X$179.4X3.3X3.8X2.6X2.
3X2.2X4.2X3.2X$181.3X2.3X3.2X7.3X2.2X2.3X2.2X4.2X3.2X$182.3X.3X3.2X7.
2X3.2X2.3X2.2X4.2X3.2X$178.2X2.3X.3X3.3X3.X2.2X2.3X2.3X2.2X4.2X3.2X$
179.5X3.4X2.6X3.6X2.3X2.4X2.2X3.2X$178.X7.X5.X8.X7.X10.X$175.65X4$
176.2X11.2X17.2X$176.2X11.2X17.2X$176.2X11.2X17.2X$176.2X11.2X17.2X$
176.2X3.5X3.7X4.5X3.2X2.5X$176.2X2.2X2.3X2.3X2.3X2.2X2.3X2.2X2.2X2.X$
176.2X7.2X2.2X4.2X.2X4.2X2.2X2.2X$176.2X3.6X2.2X4.2X.8X2.2X2.4X$176.
2X2.2X3.2X2.2X4.2X.2X8.2X4.3X$176.2X2.2X3.2X2.2X4.2X.3X7.2X5.3X$176.
2X2.2X2.3X2.3X2.3X2.2X3.2X2.2X.2X2.3X$176.2X3.6X2.7X4.5X3.2X2.5X$175.
X4.X7.X10.X7.X3.X$174.46X4$174.46V3$174.46T3$174.46R3$174.46P3$174.
46N3$174.46L3$174.46J3$174.46H3$174.46D3$174.46B3$187.2F$188.F$185.3F
$185.F!

Also I looked at the LifeViewer [R]History algo. Basically it requires that the marked states are constant (and modified at display time based on the underlying cell being dead or alive) and do not change over generations which is why the 5->4->5 final transition goes back to 5 rather than to 3.

I can switch it to use a variant of the [R]Super algo:
pros:
1) it will [be] consistent with Golly
2) it will actually simplify a fair amount of code (both algo and rendering)
cons:
1) it will be up to 10x slower in the extreme cases simulating [R]History patterns
(note: [R]Super is also up to 10x slower in the extreme cases than the current [R]History implementation)

If it simplifies some code, I'd vote for making the change, just for the sake of consistency. Yellow state 5 is hardly ever actually used, and it could usually be replaced with state 3 in any rare weird cases where performance is an issue.

Any time you're ready to put an experimental version up on the forums and wiki, I can jump in and explain the change on the LifeSuper thread.

[rowett]

The only place on the color wheel that doesn't seem overused is the white-to-black continuum -- we just have white, dark gray, and black. How about if state #20 is just something like silver, (192 192 192)?

Yes, much better.

If it simplifies some code, I'd vote for making the change, just for the sake of consistency. Yellow state 5 is hardly ever actually used, and it could usually be replaced with state 3 in any rare weird cases where performance is an issue.

The performance is not due to state 5. It's due to the [R]Super style [R]History algo being much slower than LifeViewer's current one. Oh and neither [R]Super or (the potential new [R]History) support fading history yet.

Any time you're ready to put an experimental version up on the forums and wiki, I can jump in and explain the change on the LifeSuper thread.

Done, I've released Build 566 with the new colours. I've also changed them in Golly so I don't forget before the release. If they need updating again, no problem.

[rowett]

Build 566 is now available on the lazyslug server

Enhancements in this build:

• new [R]Super colours
• abbreviate [R]Super state names in Help->Info->Theme

[dvgrn]

If it simplifies some code, I'd vote for making the change, just for the sake of consistency. Yellow state 5 is hardly ever actually used, and it could usually be replaced with state 3 in any rare weird cases where performance is an issue.

The performance is not due to state 5. It's due to the [R]Super style [R]History algo being much slower than LifeViewer's current one. Oh and neither [R]Super or (the potential new [R]History) support fading history yet.

Ah, in that case the choice isn't so clear I guess. I still think it's reasonable for LifeViewer to handle state 5 a little differently if it makes things run faster.

On the other hand, the LifeSuper algo may be slower than the LifeHistory one, but it's still pretty fast -- certainly much faster than the old rule-table-based support. I haven't seen a lot of cases yet where "LifeSuper speed" seems too slow. So I'd probably be happy either way.

I'll put a note on the LifeSuper thread now and see what people think of the adjusted colors.

[rowett]

Build 567 is now available on the lazyslug server

Fixes in this build:

• ensure random fill% stays in valid range

[rowett]

I saw some requests for population details in selections. I'm planning to add to LifeViewer at some point. I've added a prototype to Golly (not for the current release). Is this useful or feature bloat?

selpop.png (185.45 KiB) Viewed 91 times

[dvgrn]

I saw some requests for population details in selections. I'm planning to add to LifeViewer at some point. I've added a prototype to Golly (not for the current release). Is this useful or feature bloat?

I would definitely use it from time to time, e.g. for finding the minimum population of an oscillator or the number of gliders in a synthesis recipe.

Especially now that the Super makes it easier to color gliders one color and the intermediate target objects a different color, it could make it a lot easier to double-check glider costs.