Page 24 of 178
Re: Pattern viewer for forum threads
Posted: February 2nd, 2019, 2:56 pm
by muzik
Is "Posion Theme" a typo?
Re: Pattern viewer for forum threads
Posted: February 2nd, 2019, 4:45 pm
by rowett
muzik wrote:Is "Posion Theme" a typo?
Yes, well spotted!
Re: Pattern viewer for forum threads
Posted: February 2nd, 2019, 5:25 pm
by muzik
There's a few more rule names committed to catagolue, as well as some in a merge request that haven't been accepted yet, in case you're willing to add these.
(Also, the name of
this rule, which I haven't committed yet.)
Re: Pattern viewer for forum threads
Posted: February 2nd, 2019, 6:31 pm
by rowett
muzik wrote:There's a few more rule names committed to catagolue
Thanks, I added the committed ones.
Re: Pattern viewer for forum threads
Posted: February 2nd, 2019, 6:34 pm
by rowett
muzik wrote:On the "splitting the colours further" thing I mentioned earlier
Yep looks good. What we need for each Theme is a 2-state definition (which we have except for "Generations") and a multi-state definition (which we only have for "Generations").
The per-Theme grid line colour and interval is now implemented. I decided against setting grid lines on or off in a Theme - it just changes the grid line colours and interval. Also script commands will override the Theme settings.
Re: Pattern viewer for forum threads
Posted: February 2nd, 2019, 7:30 pm
by muzik
The specification for Generations with 2 states (and blues multistate) should be mentioned in the post; I'll have a crack at making some generations definitions for the remaining themes.
Also, can the multistate Larger than Life aliases be separated from the 2-state rules? They do use the same notation, but to me it's a bit like lumping outer-totalistic rules with Generations.
The alias string for Pedestrian Life without Death also intersects the rulestrng; might be worth looking into.
Re: Pattern viewer for forum threads
Posted: February 2nd, 2019, 8:29 pm
by muzik
Here's two more theme suggestions (run with the latest LifeViewer, the forum build doesn't work with these):
Golly Generations
Code: Select all
x = 9, y = 6, rule = /2/3
4.A$2.AB.BA$AB2.A2.BA$AB2.A2.BA$2.AB.BA$4.A!
[[ COLOR BACKGROUND 48 48 48 COLOR ALIVE 255 255 0 COLOR ALIVERAMP 255 0 0 COLOR DEAD 48 48 48 COLOR DEADRAMP 48 48 48 COLOR GRID 80 80 80 COLOR GRIDMAJOR 112 112 112 GRID GRIDMAJOR 10 ]]
Golly LtL Generations
Code: Select all
x = 2, y = 2, rule = R10,C3,M1,S2..3,B3..3,NM
2A$.A!
[[ COLOR BACKGROUND 48 48 48 COLOR ALIVE 255 0 0 COLOR ALIVERAMP 255 255 0 COLOR DEAD 48 48 48 COLOR DEADRAMP 48 48 48 COLOR GRID 80 80 80 COLOR GRIDMAJOR 112 112 112 GRID GRIDMAJOR 10 ]]
EDIT: there seems to be a bug with grid colours: after pasting one of these into the viewer and loading it, try then switching to the mcell theme. the grid lines are obviously the wrong colour.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 3:30 am
by rowett
muzik wrote:Also, can the multistate Larger than Life aliases be separated from the 2-state rules? They do use the same notation, but to me it's a bit like lumping outer-totalistic rules with Generations.
Yes
muzik wrote:The alias string for Pedestrian Life without Death also intersects the rulestrng; might be worth looking into.
Help information now has improved text wrapping.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 3:31 am
by rowett
muzik wrote:there seems to be a bug with grid colours: after pasting one of these into the viewer and loading it, try then switching to the mcell theme. the grid lines are obviously the wrong colour.
It's not a bug it's just not finished yet. At the moment user defined grid lines settings overwrite all themes. In the future they will just impact the Custom theme.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 9:00 am
by muzik
Outer-totalistic Generations and isotropic non-totalistic Generations should probably also be split. I'd recommend keeping the lists named and ordered similarly to how they are on the Rules page on Catagolue.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 9:16 am
by muzik
Here's some rough ideas: (ALIVE would be identical to DYING in all cases)
Poison Theme
GENDEADRAMP 0 64 0
DYING 0 255 255
DYINGRAMP 0 128 0
Grey Theme
GENDEADRAMP 52 52 52
DYING 176 176 176
DYINGRAMP 104 104 104
Inverse Theme
GENDEADRAMP 0 0 0
DYING 0 0 0
DYINGRAMP 255 255 255
(It's a bit difficult to visualize and translate these when the variables aren't actually implemented yet...)
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 9:51 am
by muzik
New aliases have been committed to Catagolue.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 9:59 am
by rowett
muzik wrote:Outer-totalistic Generations and isotropic non-totalistic Generations should probably also be split. I'd recommend keeping the lists named and ordered similarly to how they are on the Rules page on Catagolue.
Done
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 10:00 am
by rowett
rowett wrote:muzik wrote:there seems to be a bug with grid colours: after pasting one of these into the viewer and loading it, try then switching to the mcell theme. the grid lines are obviously the wrong colour.
It's not a bug it's just not finished yet. At the moment user defined grid lines settings overwrite all themes. In the future they will just impact the Custom theme.
Now when any custom element is defined with script commands it triggers the creation of the Custom Theme which will then be the default.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 10:08 am
by rowett
muzik wrote:New aliases have been committed to Catagolue.
Added, thanks.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 10:51 am
by rowett
muzik wrote:On the "splitting the colours further" thing I mentioned earlier
For 2-state rules I think I have DEAD and DEADRAMP the wrong way round.
A newly dead cell should be in DEAD colour and then move to DEADRAMP colour over HISTORYSTATES generations. So the lifecycle from nothing -> alive for many generations -> dead for many generations becomes: BACKGROUND, ALIVE -> ALIVERAMP, DEAD ->DEADRAMP.
For multi-state rules we have the following colours:
- ALIVE - for cells just born
- DYING - for cells just starting dying
- DYINGRAMP - for cells one generation from death
- DEAD - for cells just died
- DEADRAMP - for cells dead a long time
- BACKGROUND - for cells never alive
The lifecycle from nothing -> alive for many generations -> dying for several generations -> dead for HISTORYSTATES generations becomes: BACKGROUND, ALIVE, DYING -> DYINGRAMP, DEAD -> DEADRAMP.
I'll let you know when it's implemented.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 11:52 am
by muzik
One more thing: since genext support is on the backlog, it might be a good idea to define one more colour setting (BABY) which the rulespace would use for newly born cells that then transition into ALIVE cells. For the Generations Theme I'd propose that this colour would be set to green 0 255 0, since yellow is halfway between green and red in RGB terms.
Of course, this colour would go unused until the rulespaces using it are actually implemented.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 1:46 pm
by muzik
Will the Golly LtL theme be added? (It'd also be easy to fuse it with the LtL Generations theme, given the new colour settings.)
There's also a few more rule names committed on Catagolue.
Re: Pattern viewer for forum threads
Posted: February 3rd, 2019, 6:11 pm
by muzik
I noticed that LowDeath and LowFlockDeath were both assigned to the same rulestring. This turned out to be my fault, since I committed the rules to Catagolue and accidentally made the assigned rulestrings identical. I've since committed a fix (which has now since been implemented). This fix comes with even more rulestring aliases which you can add if desired.
Re: Pattern viewer for forum threads
Posted: February 4th, 2019, 4:43 am
by rowett
muzik wrote:Will the Golly LtL theme be added? (It'd also be easy to fuse it with the LtL Generations theme, given the new colour settings.)
The multi-state version of the Golly theme will look like Golly's LtL colours (i.e. yellow -> red, dark grey background, no history). That will be used for Generations, HROT and LtL patterns. I'm not planning on implementing Golly's (red -> yellow) Generations colours since I think they should change to match LtL.
Re: Pattern viewer for forum threads
Posted: February 4th, 2019, 10:57 am
by rowett
muzik wrote:I noticed that LowDeath and LowFlockDeath were both assigned to the same rulestring. This fix comes with even more rulestring aliases which you can add if desired.
Added and fixed, thanks.
Re: Pattern viewer for forum threads
Posted: February 4th, 2019, 10:58 am
by rowett
rowett wrote:muzik wrote:On the "splitting the colours further" thing I mentioned earlier
I'll let you know when it's implemented.
This is now implemented and I've done a first pass at the multi-state Themes.
Re: Pattern viewer for forum threads
Posted: February 4th, 2019, 12:03 pm
by muzik
Caterer and Catagolue do have theme colours for Generations-type rules. Here's an example for Catagolue's:
https://catagolue.appspot.com/object/xq ... g10b2s0345
As for MCell, it appears that all 255 state colours are predefined and not automatically generated on the fly, so we'd have to set new definitions to make it simulate Generations patterns with its colour scheme, which is probably not a priority.
Re: Pattern viewer for forum threads
Posted: February 4th, 2019, 12:21 pm
by muzik
Some generations themes have the colour -1 -1 -1 assigned to them, which I suspect causes some cell states to appear completely white (see the genhistory page).
EDIT: seems there's a lot,weirder stuff going on with the patterns on that page...
Re: Pattern viewer for forum threads
Posted: February 4th, 2019, 12:36 pm
by rowett
muzik wrote:Some generations themes have the colour -1 -1 -1 assigned to them, which I suspect causes some cell states to appear completely white (see the genhistory page).
EDIT: seems there's a lot,weirder stuff going on with the patterns on that page...
This should be fixed.