lifeviewer bug

Has something gone haywire? Let us know about it!
hotdogPi
Posts: 1587
Joined: August 12th, 2020, 8:22 pm

Re: lifeviewer bug

Post by hotdogPi » November 30th, 2022, 9:41 pm

AlbertArmStain wrote:
November 30th, 2022, 5:00 pm
pressing "enter" on mac doesn't work
Works for me on Safari.
User:HotdogPi/My discoveries

Periods discovered: 5-16,⑱,⑳G,㉑G,㉒㉔㉕,㉗-㉛,㉜SG,㉞㉟㊱㊳㊵㊷㊹㊺㊽㊿,54G,55G,56,57G,60,62-66,68,70,73,74S,75,76S,80,84,88,90,96
100,02S,06,08,10,12,14G,16,17G,20,26G,28,38,47,48,54,56,72,74,80,92,96S
217,486,576

S: SKOP
G: gun

Sokwe
Moderator
Posts: 2645
Joined: July 9th, 2009, 2:44 pm

Re: lifeviewer bug

Post by Sokwe » December 4th, 2022, 9:55 pm

Setting the WIDTH to values lower than 600 seems to give a thumbnail of WIDTH 600. Values above 600 seem to work properly. For example, below are thumbnails of WIDTH 480, 600, and 620 respectively. The first two appear to be the same size, while the third is slightly wider.

Code: Select all

#C [[ THEME 6 GRID GRIDMAJOR 0 SUPPRESS THUMBLAUNCH ]]
#C [[ THUMBSIZE 4 HEIGHT 320 WIDTH 480 ]]
#C [[ THUMBZOOM 64 ]]
x = 6, y = 3, rule = B3/S23
3o$o2bo$o2bobo!

Code: Select all

#C [[ THEME 6 GRID GRIDMAJOR 0 SUPPRESS THUMBLAUNCH ]]
#C [[ THUMBSIZE 4 HEIGHT 320 WIDTH 600 ]]
#C [[ THUMBZOOM 64 ]]
x = 6, y = 3, rule = B3/S23
3o$o2bo$o2bobo!

Code: Select all

#C [[ THEME 6 GRID GRIDMAJOR 0 THUMBLAUNCH ]]
#C [[ THUMBSIZE 4 HEIGHT 320 WIDTH 620 ]]
#C [[ THUMBZOOM 64 ]]
x = 6, y = 3, rule = B3/S23
3o$o2bo$o2bobo!
Another issue is that toroidal universes don't obey my placement of the pattern on the torus. For example, the following rle places the leftmost vertical line at the left edge (edit: actually the 4th column from the left) of the bounded area, but LifeViewer places the leftmost vertical line on the 6th column from the left.

Code: Select all

x = 12, y = 5, rule = B3/S23:T12,5
3bo4bo$3bo4bo$3bo4bo$3bo4bo$3bo4bo!
The following patterns have different placements of the vertical lines (leftmost vertical line in the 2nd and 3rd column from the left respectively), but they appear identical in the viewer:

Code: Select all

x = 12, y = 5, rule = B3/S23:T12,5
bo4bo$bo4bo$bo4bo$bo4bo$bo4bo!

Code: Select all

x = 12, y = 5, rule = B3/S23:T12,5
2bo4bo$2bo4bo$2bo4bo$2bo4bo$2bo4bo!
-Matthias Merzenich

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 5th, 2022, 8:55 am

Sokwe wrote:
December 4th, 2022, 9:55 pm
Setting the WIDTH to values lower than 600 seems to give a thumbnail of WIDTH 600. Values above 600 seem to work properly.
Early versions of LifeViewer allowed the WIDTH to be as small as 480. This was later changed to 600. To avoid breaking earlier patterns LifeViewer will accept (but silently change to 600) WIDTH values from 480 to 599.
Sokwe wrote:
December 4th, 2022, 9:55 pm
Another issue is that toroidal universes don't obey my placement of the pattern on the torus.
Fixed, thanks for reporting!

Sokwe
Moderator
Posts: 2645
Joined: July 9th, 2009, 2:44 pm

Re: lifeviewer bug

Post by Sokwe » December 6th, 2022, 8:38 am

The following pattern has a label placed at (2, 0.5) as measured from the top left corner of the finite bounded area. However, if I hover over the label in the viewer, it says the coordinates are (-4, -0.5) (Edit: actually (-4, -1.5)).

Code: Select all

#C [[ THEME 6 GRID GRIDMAJOR 0 ]]
#C [[ THUMBSIZE 4 ]]
#C [[ THUMBZOOM 64 WIDTH 768 HEIGHT 256 ]]
#C [[ X 0 Y 0 ]]
#C [[ STARTFROM 0 ]]
#C [[ COLOR LABEL Gray ]]
#C [[ LABELZOOMRANGE -2048 2048 ]]
#C [[ LABELT 0 0 0 LABEL 2 0.5 12 "0" ]]
#C [[ LABELT 1 1 0 LABEL 2 0.5 12 "1" ]]
#C [[ LABELT 2 2 0 LABEL 2 0.5 12 "2" ]]
#C [[ LABELT 3 3 0 LABEL 2 0.5 12 "3" ]]
#C [[ LABELT 4 4 0 LABEL 2 0.5 12 "4" ]]
#C [[ LABELT 5 5 0 LABEL 2 0.5 12 "5" ]]
#C [[ LABELT 6 6 0 LABEL 2 0.5 12 "6" ]]
#C [[ LABELT 7 7 0 LABEL 2 0.5 12 "7" ]]
#C [[ LABELT 8 8 0 LABEL 2 0.5 12 "8" ]]
#C [[ LABELT 9 9 0 LABEL 2 0.5 12 "9" ]]
#C [[ LABELT 10 10 0 LABEL 1.6 0.5 12 "10" ]]
#C [[ LABELT 11 11 0 LABEL 1.6 0.5 12 "11" ]]
#C [[ LOOP 12 ]]
x = 12, y = 4, rule = B3/S23:T12,4
5bo$5bo$b2o2bo$b2o2bo!
-Matthias Merzenich

AlbertArmStain
Posts: 1233
Joined: January 28th, 2022, 7:18 pm
Location: Planet Z

Re: lifeviewer bug

Post by AlbertArmStain » December 8th, 2022, 8:42 pm

run for some time, then pause and turn on “kill gliders”, and then reset. The glider that was light blue is now bluer
Run again then press “step back”, and the blue glider tries to move.

Code: Select all

 x = 25, y = 25, rule = B2ikn34-r5-n678/S0234-k5678
22b3o$24bo$23bo20$bo$3o$bo!

Sokwe
Moderator
Posts: 2645
Joined: July 9th, 2009, 2:44 pm

Re: lifeviewer bug

Post by Sokwe » December 10th, 2022, 8:49 am

The following toroidal immigration pattern doesn't show up in the viewer, being instead replaced by a single dot:

Code: Select all

x = 60, y = 8, rule = immigration:T60,8
60A2$3A2.3A2.3A2.3A2.3A2.3B2.3A2.3A2.3A2.3A2.3A2.3A$2.2A3.2A3.2A3.2A
3.2A3.2B3.2A3.2A3.2A3.2A3.2A3.2A$3A2.3A2.3A2.3A2.3A2.3B2.3A2.3A2.3A2.
3A2.3A2.3A2$60A!
If thumbnails are used, the thumbnail has the single dot, but the pop up viewer gives the correct pattern:

Code: Select all

#C [[ THUMBLAUNCH ]]
x = 60, y = 8, rule = immigration:T60,8
60A2$3A2.3A2.3A2.3A2.3A2.3B2.3A2.3A2.3A2.3A2.3A2.3A$2.2A3.2A3.2A3.2A
3.2A3.2B3.2A3.2A3.2A3.2A3.2A3.2A$3A2.3A2.3A2.3A2.3A2.3B2.3A2.3A2.3A2.
3A2.3A2.3A2$60A!
Launching from a code box works fine:

Code: Select all

x = 60, y = 8, rule = immigration:T60,8
60A2$3A2.3A2.3A2.3A2.3A2.3B2.3A2.3A2.3A2.3A2.3A2.3A$2.2A3.2A3.2A3.2A
3.2A3.2B3.2A3.2A3.2A3.2A3.2A3.2A$3A2.3A2.3A2.3A2.3A2.3B2.3A2.3A2.3A2.
3A2.3A2.3A2$60A!
An unrelated and very minor/low-priority issue:

When I make a white polygon with alpha 0.5 on a white background I would think I'd get a white polygon (white-on-white should still be white with transparency, right?). Instead, I get a gray polygon:

Code: Select all

#C [[ THEME 6 ]]
#C [[ POLYZOOMRANGE -2048 2048 ]]
#C [[ COLOR POLY White ]]
#C [[ POLYALPHA 0.5 ]]
#C [[ POLYFILL 0 0 0 1 1 1 1 0 0 0 64 ]]
!
Is this a bug? Intended behavior? I don't really know anything about alpha compositing so maybe this is expected. I also don't have a very meaningful need to have this "fixed", so consider it low-priority.

Finally, not a bug, but since I'm here I thought I'd ask: is there a command to remove the drop shadow on text and polygons?

Edit: Setting a theme seems to prevent the grid/gridmajor coloring from working, no matter what order I include the commands:

Code: Select all

#C [[ THEME 6 ]]
#C [[ GRID ]]
#C [[ GRIDMAJOR 10 ]]
#C [[ COLOR GRID Green ]]
#C [[ COLOR GRIDMAJOR Red ]]
!
-Matthias Merzenich

AlbertArmStain
Posts: 1233
Joined: January 28th, 2022, 7:18 pm
Location: Planet Z

Re: lifeviewer bug

Post by AlbertArmStain » December 11th, 2022, 5:47 pm

Sokwe wrote:
December 10th, 2022, 8:49 am

Launching from a code box works fine:

Code: Select all

x = 60, y = 8, rule = immigration:T60,8
60A2$3A2.3A2.3A2.3A2.3A2.3B2.3A2.3A2.3A2.3A2.3A2.3A$2.2A3.2A3.2A3.2A
3.2A3.2B3.2A3.2A3.2A3.2A3.2A3.2A$3A2.3A2.3A2.3A2.3A2.3B2.3A2.3A2.3A2.
3A2.3A2.3A2$60A!
I tried changing the rule (not the torus), and the dot showed up again.

Sokwe
Moderator
Posts: 2645
Joined: July 9th, 2009, 2:44 pm

Re: lifeviewer bug

Post by Sokwe » December 12th, 2022, 1:40 am

Sorry to post another minor, low-priority issue. The following thumbnail appears just a few pixels too thin:

Code: Select all

#C [[ THEME 6 GRID GRIDMAJOR 0 THUMBLAUNCH THUMBSIZE 4 ]]
#C [[ THUMBZOOM 60 HEIGHT 300 WIDTH 780 ]]
!
The white areas of the rightmost cells are only 12 pixels wide, instead of the expected 14 pixels. Increasing WIDTH doesn't change the width of the thumbnail until WIDTH 800, at which point it's too wide. A thumbnail with THUMBZOOM 64 WIDTH 832 seems to have the expected 15-pixel white area width for all cells.

This all might be intended behavior. Perhaps the width of the thumbnail only goes up in increments of which my chosen width is not an exact multiple. My recent issues all come from my attempts to get a very particular look for the thumbnails on the LifeWiki Lifeline pages.
-Matthias Merzenich

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 12th, 2022, 5:53 am

Sokwe wrote:
December 12th, 2022, 1:40 am
This all might be intended behavior. Perhaps the width of the thumbnail only goes up in increments of which my chosen width is not an exact multiple.
Viewers are always a multiple of 8 pixels wide.

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 12th, 2022, 8:20 am

AlbertArmStain wrote:
December 8th, 2022, 8:42 pm
run for some time, then pause and turn on “kill gliders”, and then reset. The glider that was light blue is now bluer
Fixed, thanks!

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 12th, 2022, 8:26 am

Sokwe wrote:
December 10th, 2022, 8:49 am
The following toroidal immigration pattern doesn't show up in the viewer, being instead replaced by a single dot:
Fixed, thanks!
Sokwe wrote:
December 10th, 2022, 8:49 am
When I make a white polygon with alpha 0.5 on a white background I would think I'd get a white polygon (white-on-white should still be white with transparency, right?). Instead, I get a gray polygon:
It's because the polygon has a shadow and you're seeing the shadow.
Sokwe wrote:
December 10th, 2022, 8:49 am
Finally, not a bug, but since I'm here I thought I'd ask: is there a command to remove the drop shadow on text and polygons?
There is now. I added [[ LABELSHADOW (OFF) ]], [[ ARROWSHADOW (OFF) ]] and [[ POLYSHADOW (OFF) ]] script commands. They determine whether subsequent annotations have a shadow. Default is shadow on.

Code: Select all

x=10, y=10, rule=Life
10o$10o$10o$10o$10o$10o$10o$10o$10o$10!
[[
LABEL 0 0 16 "Shadow"
LABELSHADOW OFF
LABEL 0 4 16 "No Shadow"
LABELSHADOW
LABEL 0 8 16 "Shadow again"
]]
Sokwe wrote:
December 10th, 2022, 8:49 am
Setting a theme seems to prevent the grid/gridmajor coloring from working, no matter what order I include the commands:
Fixed, thanks!

Sokwe
Moderator
Posts: 2645
Joined: July 9th, 2009, 2:44 pm

Re: lifeviewer bug

Post by Sokwe » December 12th, 2022, 5:19 pm

rowett wrote:
December 12th, 2022, 8:26 am
I added [[ LABELSHADOW (OFF) ]], [[ ARROWSHADOW (OFF) ]] and [[ POLYSHADOW (OFF) ]] script commands.
Thanks! However, on labels split by newlines only the last line has the shadow turned off:

Code: Select all

#C [[ THEME 6 ]]
#C [[ LABELSHADOW OFF ]] 
#C [[ COLOR LABEL Gray ]]
#C [[ ZOOM 16 ]]
#C [[ LABEL -5 0 10 "Line 1\nLine 2" ]]
#C [[ LABEL 5 0 10 "Line 1\nLine 2\n Line 3" ]]
!
-Matthias Merzenich

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 12th, 2022, 6:12 pm

Sokwe wrote:
December 12th, 2022, 5:19 pm
on labels split by newlines only the last line has the shadow turned off
Fixed, thanks!

AlbertArmStain
Posts: 1233
Joined: January 28th, 2022, 7:18 pm
Location: Planet Z

Re: lifeviewer bug

Post by AlbertArmStain » December 12th, 2022, 6:53 pm

Run this for awhile (the bigger the amount of generations, the better), then turn on “kill gliders”
Do one step forward to kill the glider, then start pressing “step backward”. You’ll eventually see the glider trail shorten by a lot.

Code: Select all

x = 7, y = 6, rule = B3/S23
5bo$5bo$4bobo$5bo$5bo$3o!

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 13th, 2022, 1:46 am

AlbertArmStain wrote:
December 12th, 2022, 6:53 pm
Do one step forward to kill the glider, then start pressing “step backward”. You’ll eventually see the glider trail shorten by a lot.
It's related to how the step backward feature works and won't be fixed.

Sokwe
Moderator
Posts: 2645
Joined: July 9th, 2009, 2:44 pm

Re: lifeviewer bug

Post by Sokwe » December 13th, 2022, 3:44 am

Some more oddities to report:
  • At generations that are a multiple of 64 all alive cells are white when I expect some to be cyan:

    Code: Select all

    #C [[ AGESTATES 1 ]]
    x = 49, y = 42, rule = B3/S23
    17b2o$18bo$17bo$17b4o$20bo$15b3o$4b2o8bo2bo$4bobo2b2o3b2o$bo4bo3bo$b5o
    b3o$5bobo$b3obobo$o2bob2o$2o2$19b2o$4b2o12bobo22bo$4b2o11b2o22b3o$18bo
    bo19bo$19b2o19b2o3$7b2o19b2o$8bo19bobo$5b3o22b2o11b2o$5bo22bobo12b2o$
    28b2o2$47b2o$42b2obo2bo$41bobob3o$41bobo$39b3ob5o$38bo3bo4bo$33b2o3b2o
    2bobo$31bo2bo8b2o$31b3o$28bo$28b4o$31bo$30bo$30b2o!
    Also, pressing reset makes all alive cells white rather than their original cyan. Pressing reset followed by step forward followed by step backward makes all alive cells cyan.
  • Setting a theme seems to prevent subsequent cell state coloring from working:

    Code: Select all

    #C [[ THEME 6 ]]
    #C [[ COLOR BACKGROUND Blue ]]
    #C [[ COLOR ALIVE White ]]
    #C [[ COLOR DEAD Orange ]]
    #C [[ COLOR DEADRAMP Gray ]]
    o$boo!
    This seems similar to the issue where grid colors couldn't be changed when a theme was set. There may be other things whose color can't be changed when a theme is set.
-Matthias Merzenich

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 13th, 2022, 3:50 am

Sokwe wrote:
December 6th, 2022, 8:38 am
The following pattern has a label placed at (2, 0.5) as measured from the top left corner of the finite bounded area. However, if I hover over the label in the viewer, it says the coordinates are (-4, -0.5).
Fixed, thanks!

Code: Select all

#C [[ THEME 6 GRID GRIDMAJOR 0 ]]
#C [[ THUMBSIZE 4 ]]
#C [[ THUMBZOOM 64 WIDTH 768 HEIGHT 256 ]]
#C [[ X 0 Y 0 ]]
#C [[ STARTFROM 0 ]]
#C [[ COLOR LABEL Gray ]]
#C [[ LABELZOOMRANGE -2048 2048 ]]
#C [[ LABELT 0 0 0 LABEL -4 -0.5 12 "0" ]]
#C [[ LABELT 1 1 0 LABEL -4 -0.5 12 "1" ]]
#C [[ LABELT 2 2 0 LABEL -4 -0.5 12 "2" ]]
#C [[ LABELT 3 3 0 LABEL -4 -0.5 12 "3" ]]
#C [[ LABELT 4 4 0 LABEL -4 -0.5 12 "4" ]]
#C [[ LABELT 5 5 0 LABEL -4 -0.5 12 "5" ]]
#C [[ LABELT 6 6 0 LABEL -4 -0.5 12 "6" ]]
#C [[ LABELT 7 7 0 LABEL -4 -0.5 12 "7" ]]
#C [[ LABELT 8 8 0 LABEL -4 -0.5 12 "8" ]]
#C [[ LABELT 9 9 0 LABEL -4 -0.5 12 "9" ]]
#C [[ LABELT 10 10 0 LABEL -4.4 -0.5 12 "10" ]]
#C [[ LABELT 11 11 0 LABEL -4.4 -0.5 12 "11" ]]
#C [[ LOOP 12 ]]
x = 12, y = 4, rule = B3/S23:T12,4
5bo$5bo$b2o2bo$b2o2bo!

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 16th, 2022, 2:03 am

Sokwe wrote:
December 13th, 2022, 3:44 am
At generations that are a multiple of 64 all alive cells are white when I expect some to be cyan:
Thanks for the report. It's on the backlog to fix.
Sokwe wrote:
December 13th, 2022, 3:44 am
Setting a theme seems to prevent subsequent cell state coloring from working
Fixed, thanks!

AlbertArmStain
Posts: 1233
Joined: January 28th, 2022, 7:18 pm
Location: Planet Z

Re: lifeviewer bug

Post by AlbertArmStain » December 16th, 2022, 5:29 pm

Go to this post and click the first code box.
The box shrinks by an observable amount, it also brings you to the very top of the page.
Something simular happens here by clicking the third code box.

it actually happens every time the pattern is too big

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 17th, 2022, 1:03 am

AlbertArmStain wrote:
December 16th, 2022, 5:29 pm
Go to this post and click the first code box.
The box shrinks by an observable amount, it also brings you to the very top of the page.
Something simular happens here by clicking the third code box.

it actually happens every time the pattern is too big
Fixed, thanks!

AlbertArmStain
Posts: 1233
Joined: January 28th, 2022, 7:18 pm
Location: Planet Z

Re: lifeviewer bug

Post by AlbertArmStain » December 17th, 2022, 11:17 am

The select feature has a weird offset now

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 17th, 2022, 3:21 pm

AlbertArmStain wrote:
December 17th, 2022, 11:17 am
The select feature has a weird offset now
Yes it's currently broken. WIll be fixed when I get time.

AlbertArmStain
Posts: 1233
Joined: January 28th, 2022, 7:18 pm
Location: Planet Z

Re: lifeviewer bug

Post by AlbertArmStain » December 20th, 2022, 10:57 am

rowett wrote:
December 17th, 2022, 3:21 pm
Yes it's currently broken. WIll be fixed when I get time.
Ok, thanks. Please tell me when it’s fixed.

User avatar
rowett
Moderator
Posts: 3776
Joined: January 31st, 2013, 2:34 am
Location: UK
Contact:

Re: lifeviewer bug

Post by rowett » December 21st, 2022, 5:05 am

AlbertArmStain wrote:
December 20th, 2022, 10:57 am
rowett wrote:
December 17th, 2022, 3:21 pm
Yes it's currently broken. WIll be fixed when I get time.
Ok, thanks. Please tell me when it’s fixed.
Most of the selection issues should be fixed now.

Selections are still broken for patterns in Bounded Grids where the actual pattern size is different than the specified size.

User avatar
wirehead
Posts: 245
Joined: June 18th, 2022, 2:37 pm
Location: Under a thinking cap
Contact:

Re: lifeviewer bug

Post by wirehead » December 29th, 2022, 10:22 am

On large p2 patterns (like, say, unsynthesizable oscillator 1) and speed slider set to 1x, frames are randomly dropped and the pattern appears to oscillate erratically. Are you using setTimeout for the drawing loop? Maybe requestAnimationFrame would work better.
Langton's ant: Can't play the drums, can be taught.

Post Reply