David Bell's engineless caterpillar idea revisited

For discussion of specific patterns or specific families of patterns, both newly-discovered and well-known.
User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 22nd, 2015, 1:03 am

dvgrn wrote:chris_c's search script currently only keeps the first recipe found for each object
Well I've written a script a while ago, and it keeps all the recipes per object (I reached depth of 8 in two weeks and didn't go to depth 9 which would take 3 month). For that recipe list I've written a small search utility that search combination of recipes while keeping it from getting collided into other object present in the initial golly state (thus really fine tune the placement recipe to fit exactly the specific problem on hand).

I've posted it in (the name is search <depth associated integer> + SL coordinates):
https://github.com/simsim314/Glue

The search script is called RecipeStapler.py, and you'll need to fix the path there to your Glue recipes location (see line 378 and on).

I've no tutorial on the usage, and this is actually work in progress. It's currently capable of combining two recipes together, and written well enough for me to understand it after a while. I guess you have your own ways to combine the recipes, but I can help with RecipeStapler if needed (if nothing else it has a feature to place all recipes from recipe list).

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 22nd, 2015, 8:31 am

simsim314 wrote:I've no tutorial on the usage, and this is actually work in progress. It's currently capable of combining two recipes together, and written well enough for me to understand it after a while. I guess you have your own ways to combine the recipes, but I can help with RecipeStapler if needed (if nothing else it has a feature to place all recipes from recipe list).
Thanks! Looks straightforward enough. I posted exactly the same functionality (with just about the same level of usability) for P2 slow recipes, which needed to be stapled together for the Geminoid linear propagator and the spiral-growth pattern. My stapler script was posted here.

Other differences besides P2 vs. P1:
  • my recipe lists use Paul Chapman's lane-numbering system, so it's possible to negate every number and get a mirror-image recipe;
  • all my recipes are labeled by still-life name instead of lists of coordinates;
  • recipes for constellations of still lifes are included, as well as clean single objects (this came in handy a lot);
  • my depth-8 search wasn't exhaustive, because I cheated and re-used an existing depth-9 search for block moves. Truncating depth-9 recipes always gives well-behaved P2 constellations, of course, but there are a lot more recipes out there that can't be completed into a block move...!
Looks like you've gotten several steps farther along than I did. If I have time, I'll try putting together a few MWSS recipes using your system.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 22nd, 2015, 3:01 pm

dvgrn wrote:my recipe lists use Paul Chapman's lane-numbering system
Well this is arguably more intuitive for slow salvo. But placing the block at (0, 0) and glider at (10, 10) and calling this lane 0, is more intuitive as general approach.

The main difference that you can use -x, and my recipes need to use 1 - x. Anyway I stored both recipes anyway.
dvgrn wrote:all my recipes are labeled by still-life name instead of lists of coordinates
List of coordinates was used by the glue script, to search for any SL, I wasn't limited to predefined SL list.
dvgrn wrote:recipes for constellations of still lifes are included, as well as clean single objects
This one I really missed, and this caused me to manually alter almost any recipe, to avoid deleting a block, and use it for the next recipe.
dvgrn wrote:my depth-8 search wasn't exhaustive
Well p2 branching factor is much higher so I'm not sure the comparison is fair.
dvgrn wrote: My stapler script was posted here.
The major advantage of my script, is that it "empirically" checks if the recipe is valid. You can add initial state into golly, and then search for recipe - the request is not to ruin the initial state. So if for example I want a recipe that doesn't touch the *WSS lane, I just place on that lane a bunch of blocks and the search will know not to come close to that lane. Or for example if I want to add some tight SL near the lane and some other SL, the script is well suited for the task.

My recipe list advantage is that it contains all the possible (even if almost equivalent) recipes. Think that for some reason the deletion glider should be of some specific lane, because there are some SL above and below. My stapler will know to use the exact recipe that will work there on that specific lane, and the recipe list will have this specific recipe as well.

EDIT
dvgrn wrote: I'll try putting together a few MWSS recipes using your system.
Thx any help will be much appreciated.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 22nd, 2015, 8:12 pm

Here is combined recipe of HWSS with one of the adjustment recipes (block -> HWSS + block):

Code: Select all

x = 5459, y = 5432, rule = B3/S23
2o$2o75$80b3o$80bo$81bo124$208b3o$208bo$209bo124$336b3o$336bo$337bo
125$464b3o$464bo$465bo118$592b3o$592bo$593bo108$720b3o$720bo$721bo124$
848b3o$848bo$849bo124$976b3o$976bo$977bo118$1104b3o$1104bo$1105bo113$
1232b3o$1232bo$1233bo131$1360b3o$1360bo$1361bo144$1488b3o$1488bo$1489b
o128$1616b3o$1616bo$1617bo134$1744b3o$1744bo$1745bo112$1872b3o$1872bo$
1873bo119$2000b3o$2000bo$2001bo133$2128b3o$2128bo$2129bo127$2256b3o$
2256bo$2257bo136$2384b3o$2384bo$2385bo135$2512b3o$2512bo$2513bo126$
2640b3o$2640bo$2641bo124$2768b3o$2768bo$2769bo138$2896b3o$2896bo$2897b
o118$3024b3o$3024bo$3025bo114$3152b3o$3152bo$3153bo152$3280b3o$3280bo$
3281bo124$3408b3o$3408bo$3409bo124$3536b3o$3536bo$3537bo125$3664b3o$
3664bo$3665bo118$3792b3o$3792bo$3793bo110$3920b3o$3920bo$3921bo148$
4048b3o$4048bo$4049bo136$4176b3o$4176bo$4177bo102$4304b3o$4304bo$4305b
o140$4432b3o$4432bo$4433bo103$4560b3o$4560bo$4561bo117$4688b3o$4688bo$
4689bo125$4816b3o$4816bo$4817bo144$4944b3o$4944bo$4945bo127$5072b3o$
5072bo$5073bo141$5200b3o$5200bo$5201bo127$5328b3o$5328bo$5329bo109$
5456b3o$5456bo$5457bo!
And the recipe is:
[-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -12, -2, -26, -12, -35, -44, -45, -27, -26, -11, -10, -27]

BTW I've extracted the 8 adjustment recipes to list of movement commands:

adjustments= [[-4, -13, -14, 4, 5], [-4, -6, -8, -11, -3, -12], [-4, -11, -14, -5, 9, 6], [-4, -13, -8, -9, -12, -7], [-4, -13, -13, -11, -16, -4], [-4, -13, -13, -12, 4, -16], [-4, -6, -13, -10, -18, -9], [-4, 4, -10, -13, -4]]

I'm not exactly sure how they were ordered.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 22nd, 2015, 10:42 pm

simsim314 wrote:... due to some complications, I guess I'll also need to add N gliders "void" operation, to keep all the recipes at the same tick. Or maybe generate k gliders adjustment recipes for each of the adjustments, so I can "lock" the final tick.
Am I understanding correctly that this has to do with making a caterloopillar toolkit that works at any period? That is, it will be simpler if the MWSS edge-shooter seed recipes with the eight different timings all take exactly the same number of gliders, and the same for all the HWSS seed recipes, so that the eight variants are really interchangeable? (No need for HWSS and MWSS recipes to be the same size, I think.)

That seems as if it might be fairly simple to handle. At least, it's generally trivial to shoot down an extra block with one glider, two gliders... N gliders, just by doing N-1 block-pulls before the final annihilation.

But is it more complicated than just the number of gliders? Does it make a difference if some N-glider salvos take up more vertical space than other N-glider salvos? Maybe the idea should really be to have all the HWSS or MWSS variant recipes take up exactly the same amount of vertical space in the spaceship, so they can be painlessly swapped out for each other...?
simsim314 wrote:I'm leaning toward using the timing adjustment solution, it's ugly and efficient, and costs about 20% more than the optimal 3 SL for each timing.
It seems as if this might be the simplest line of research to pursue, and not really all that ugly. Find one particularly cheap recipe each for MWSS and HWSS seeds plus an extra block at a safe distance. Then come up with eight equal-length timing adjustment tricks (whatever that means exactly -- I'm not quite clear on the details yet).

Then the same timing toolkit could be used to adjust both MWSS and HWSS timing. What could be better than that? It might even be 20% more expensive than a 20%-cheaper *WSS recipe, so we'd end up a little ahead.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 23rd, 2015, 7:00 am

dvgrn wrote: it will be simpler if the MWSS edge-shooter seed recipes with the eight different timings all take exactly the same number of gliders
Well as you mention further, the "normalization" is not only for the number of glider but the parity of the SL placed. But I'm really not sure if this is necessary at all. I would start from any number of gliders for the 16 recipes.

I'm still a bit struggling to figure out the arithmetics. For now I'm trying to build c/11 caterloopillar, just to check out the arithmetical nuances on specific case, and for that case the number of gliders in recipes wasn't matter.
dvgrn wrote:Find one particularly cheap recipe each for MWSS and HWSS seeds
Well this already pretty challenging. I agree that this is not as memory intensive as finding optimized recipes for all cases, but in general if you find one such recipe finding more of the same is straightforward.
dvgrn wrote:whatever that means exactly -- I'm not quite clear on the details yet
Well the details are very simple. Because the SLs can move only in even steps to fit the *WSS salvo "reader", so to fit the exact timing and parity of generated *WSS we need output glider to be "shifted" in phase and lane. You can look at input glider as the last glider in the slow-salvo recipe, and the output is what comes out from the one of 8 recipes.

I'm not sure if some trickery can be done by using odd number of SLs to make void operation, thus switching the parity of the input glider. I'm not sure if this trick will actually switch the lane, or maybe just do the same one lane lower. Meanwhile I see that there is no point in switching the parity, not sure exactly if this universally correct, or just for c/11 caterloopillar.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 23rd, 2015, 4:08 pm

Here are recipes for all 8 adjustments + hwss (we need only 7, but I did the one that kept the state and parity by mistake so I post it as well):

Code: Select all

x = 5971, y = 12432, rule = B3/S23
2o$2o75$80b3o$80bo$81bo124$208b3o$208bo$209bo124$336b3o$336bo$337bo
125$464b3o$464bo$465bo118$592b3o$592bo$593bo108$720b3o$720bo$721bo124$
848b3o$848bo$849bo124$976b3o$976bo$977bo61$2o$2o56$1104b3o$1104bo$
1105bo17$80b3o$80bo$81bo94$1232b3o$1232bo$1233bo28$208b3o$208bo$209bo
101$1360b3o$1360bo$1361bo21$336b3o$336bo$337bo121$1488b3o$1488bo$1489b
o2$464b3o$464bo$465bo118$592b3o$592bo$593bo4$1616b3o$1616bo$1617bo102$
720b3o$720bo$721bo30$1744b3o$1744bo$1745bo92$848b3o$848bo$849bo18$
1872b3o$1872bo$1873bo104$976b3o$976bo$977bo13$2000b3o$2000bo$2001bo46$
2o$2o56$1104b3o$1104bo$1105bo17$80b3o$80bo$81bo9$2128b3o$2128bo$2129bo
83$1232b3o$1232bo$1233bo28$208b3o$208bo$209bo12$2256b3o$2256bo$2257bo
87$1360b3o$1360bo$1361bo21$336b3o$336bo$337bo24$2384b3o$2384bo$2385bo
95$1488b3o$1488bo$1489bo2$464b3o$464bo$465bo34$2512b3o$2512bo$2513bo
82$592b3o$592bo$593bo4$1616b3o$1616bo$1617bo36$2640b3o$2640bo$2641bo
64$720b3o$720bo$721bo30$1744b3o$1744bo$1745bo26$2768b3o$2768bo$2769bo
64$848b3o$848bo$849bo18$1872b3o$1872bo$1873bo52$2896b3o$2896bo$2897bo
50$976b3o$976bo$977bo13$2000b3o$2000bo$2001bo46$2o$2o4$3024b3o$3024bo$
3025bo50$1104b3o$1104bo$1105bo17$80b3o$80bo$81bo9$2128b3o$2128bo$2129b
o32$3152b3o$3152bo$3153bo49$1232b3o$1232bo$1233bo28$208b3o$208bo$209bo
12$2256b3o$2256bo$2257bo57$3280b3o$3280bo$3281bo28$1360b3o$1360bo$
1361bo21$336b3o$336bo$337bo24$2384b3o$2384bo$2385bo45$3408b3o$3408bo$
3409bo48$1488b3o$1488bo$1489bo2$464b3o$464bo$465bo34$2512b3o$2512bo$
2513bo34$3536b3o$3536bo$3537bo46$592b3o$592bo$593bo4$1616b3o$1616bo$
1617bo36$2640b3o$2640bo$2641bo33$3664b3o$3664bo$3665bo29$720b3o$720bo$
721bo30$1744b3o$1744bo$1745bo26$2768b3o$2768bo$2769bo27$3792b3o$3792bo
$3793bo35$848b3o$848bo$849bo18$1872b3o$1872bo$1873bo52$2896b3o$2896bo
1023b3o$2897bo1022bo$3921bo49$976b3o$976bo$977bo13$2000b3o$2000bo$
2001bo46$2o$2o4$3024b3o$3024bo$3025bo29$4048b3o$4048bo$4049bo19$1104b
3o$1104bo$1105bo17$80b3o$80bo$81bo9$2128b3o$2128bo$2129bo32$3152b3o$
3152bo$3153bo49$1232b3o$1232bo$1233bo2942b3o$4176bo$4177bo26$208b3o$
208bo$209bo12$2256b3o$2256bo$2257bo57$3280b3o$3280bo$3281bo$4304b3o$
4304bo$4305bo25$1360b3o$1360bo$1361bo21$336b3o$336bo$337bo24$2384b3o$
2384bo$2385bo45$3408b3o$3408bo$3409bo17$4432b3o$4432bo$4433bo29$1488b
3o$1488bo$1489bo2$464b3o$464bo$465bo34$2512b3o$2512bo$2513bo32$4560b3o
$4560bo$3536b3o1022bo$3536bo$3537bo46$592b3o$592bo$593bo4$1616b3o$
1616bo$1617bo36$2640b3o$2640bo$2641bo23$4688b3o$4688bo$4689bo8$3664b3o
$3664bo$3665bo29$720b3o$720bo$721bo30$1744b3o$1744bo$1745bo26$2768b3o$
2768bo$2769bo24$4816b3o$4816bo$4817bo$3792b3o$3792bo$3793bo35$848b3o$
848bo$849bo18$1872b3o$1872bo$1873bo52$2896b3o$2896bo1023b3o$2897bo
1022bo$3921bo29$4944b3o$4944bo$4945bo18$976b3o$976bo$977bo13$2000b3o$
2000bo$2001bo46$2o$2o4$3024b3o$3024bo$3025bo20$4048b3o$4048bo$4049bo
17$5072b3o$5072bo$5073bo9$1104b3o$1104bo$1105bo17$80b3o$80bo$81bo9$
2128b3o$2128bo$2129bo32$3152b3o$3152bo$3153bo22$4176b3o$4176bo$4177bo
25$1232b3o$1232bo$1233bo15$5200b3o$5200bo$5201bo11$208b3o$208bo$209bo
12$2256b3o$2256bo$2257bo57$3280b3o$3280bo$3281bo9$4304b3o$4304bo$4305b
o17$1360b3o$1360bo$1361bo11$5328b3o$5328bo$5329bo8$336b3o$336bo$337bo
24$2384b3o$2384bo$2385bo45$3408b3o$3408bo$3409bo23$4432b3o$4432bo$
4433bo$5456b3o$5456bo$5457bo20$1488b3o$1488bo$1489bo2$464b3o$464bo$
465bo34$2512b3o$2512bo$2513bo34$3536b3o$3536bo$3537bo19$4560b3o$4560bo
$4561bo25$592b3o$592bo$593bo4$1616b3o$1616bo$1617bo36$2640b3o$2640bo$
2641bo33$3664b3o$3664bo$3665bo29$720b3o$720bo$721bo12$4688b3o$4688bo$
4689bo16$1744b3o$1744bo$1745bo26$2768b3o$2768bo$2769bo27$3792b3o$3792b
o$3793bo35$848b3o$848bo$849bo10$4816b3o$4816bo$4817bo6$1872b3o$1872bo$
1873bo52$2896b3o$2896bo1023b3o$2897bo1022bo$3921bo49$976b3o$976bo$977b
o10$4944b3o$4944bo$4945bo$2000b3o$2000bo$2001bo46$2o$2o4$3024b3o$3024b
o$3025bo20$4048b3o$4048bo$4049bo28$1104b3o$1104bo$1105bo16$5072b3o$80b
3o4989bo$80bo4992bo$81bo9$2128b3o$2128bo$2129bo32$3152b3o$3152bo$3153b
o23$4176b3o$4176bo$4177bo24$1232b3o$1232bo$1233bo26$5200b3o$5200bo$
208b3o4990bo$208bo$209bo12$2256b3o$2256bo$2257bo57$3280b3o$3280bo$
3281bo6$4304b3o$4304bo$4305bo20$1360b3o$1360bo$1361bo21$336b3o$336bo$
337bo6$5328b3o$5328bo$5329bo16$2384b3o$2384bo$2385bo45$3408b3o$3408bo$
3409bo4$4432b3o$4432bo$4433bo42$1488b3o$1488bo$1489bo2$464b3o4989b3o$
464bo4991bo$465bo4991bo34$2512b3o$2512bo$2513bo34$3536b3o$3536bo$3537b
o9$4560b3o$4560bo$4561bo35$592b3o$592bo$593bo4$1616b3o$1616bo$1617bo$
5584b3o$5584bo$5585bo33$2640b3o$2640bo$2641bo33$3664b3o$3664bo$3665bo$
4688b3o$4688bo$4689bo26$720b3o$720bo$721bo25$5712b3o$5712bo$5713bo3$
1744b3o$1744bo$1745bo26$2768b3o$2768bo$2769bo27$3792b3o$3792bo$3793bo
2$4816b3o$4816bo$4817bo31$848b3o$848bo$849bo17$5840b3o$1872b3o3965bo$
1872bo3968bo$1873bo52$2896b3o$2896bo1023b3o$2897bo1022bo$3921bo15$
4944b3o$4944bo$4945bo32$976b3o$976bo$977bo2$5968b3o$5968bo$5969bo9$
2000b3o$2000bo$2001bo46$2o$2o4$3024b3o$3024bo$3025bo30$4048b3o$4048bo$
4049bo$5072b3o$5072bo$5073bo15$1104b3o$1104bo$1105bo17$80b3o$80bo$81bo
9$2128b3o$2128bo$2129bo32$3152b3o$3152bo$3153bo43$4176b3o$4176bo$4177b
o4$1232b3o$1232bo$1233bo8$5200b3o$5200bo$5201bo18$208b3o$208bo$209bo
12$2256b3o$2256bo$2257bo57$3280b3o$3280bo$3281bo8$4304b3o$4304bo$4305b
o18$1360b3o$1360bo$1361bo3966b3o$5328bo$5329bo19$336b3o$336bo$337bo24$
2384b3o$2384bo$2385bo45$3408b3o$3408bo$3409bo$4432b3o$4432bo$4433bo20$
5456b3o$5456bo$5457bo23$1488b3o$1488bo$1489bo2$464b3o$464bo$465bo34$
2512b3o$2512bo$2513bo34$3536b3o$3536bo$3537bo8$4560b3o$4560bo$4561bo
25$5584b3o$5584bo$5585bo9$592b3o$592bo$593bo4$1616b3o$1616bo$1617bo36$
2640b3o$2640bo$2641bo33$3664b3o$3664bo$3665bo8$4688b3o$4688bo$4689bo
19$720b3o$720bo$721bo15$5712b3o$5712bo$5713bo13$1744b3o$1744bo$1745bo
26$2768b3o$2768bo$2769bo27$3792b3o$3792bo$3793bo13$4816b3o$4816bo$
4817bo20$848b3o$848bo$849bo4990b3o$5840bo$5841bo16$1872b3o$1872bo$
1873bo52$2896b3o$2896bo1023b3o$2897bo1022bo$3921bo34$4944b3o$4944bo$
4945bo13$976b3o$976bo$977bo13$2000b3o$2000bo$2001bo51$3024b3o$3024bo$
3025bo29$4048b3o$4048bo$4049bo14$5072b3o$5072bo$5073bo3$1104b3o$1104bo
$1105bo28$2128b3o$2128bo$2129bo32$3152b3o$3152bo$3153bo47$5200b3o$
5200bo$1232b3o3966bo$1232bo$1233bo2942b3o$4176bo$4177bo40$2256b3o$
2256bo$2257bo52$4304b3o$4304bo$4305bo3$3280b3o$3280bo$3281bo28$1360b3o
$1360bo$1361bo11$5328b3o$5328bo$5329bo34$2384b3o$2384bo$2385bo45$3408b
3o$3408bo$3409bo25$4432b3o$4432bo$4433bo16$5456b3o$5456bo$5457bo3$
1488b3o$1488bo$1489bo38$2512b3o$2512bo$2513bo34$3536b3o$3536bo$3537bo
18$4560b3o$4560bo$4561bo8$5584b3o$5584bo$5585bo22$1616b3o$1616bo$1617b
o36$2640b3o$2640bo$2641bo33$3664b3o$3664bo$3665bo10$4688b3o$4688bo$
4689bo49$1744b3o$1744bo$1745bo26$2768b3o$2768bo$2769bo27$3792b3o$3792b
o$3793bo18$4816b3o$4816bo$4817bo35$1872b3o$1872bo$1873bo52$2896b3o$
2896bo1023b3o$2897bo1022bo$3921bo36$4944b3o$4944bo$4945bo26$2000b3o$
2000bo$2001bo51$3024b3o$3024bo$3025bo19$4048b3o$4048bo$4049bo19$5072b
3o$5072bo$5073bo38$2128b3o$2128bo$2129bo32$3152b3o$3152bo$3153bo30$
4176b3o$4176bo$4177bo32$5200b3o$5200bo$5201bo27$2256b3o$2256bo$2257bo
57$3280b3o$3280bo$3281bo3$4304b3o$4304bo$4305bo37$5328b3o$5328bo$5329b
o33$2384b3o$2384bo$2385bo45$3408b3o$3408bo$3409bo5$4432b3o$4432bo$
4433bo36$5456b3o$5456bo$5457bo43$2512b3o$2512bo$2513bo34$3536b3o1021b
3o$3536bo1023bo$3537bo1023bo28$5584b3o$5584bo$5585bo60$2640b3o$2640bo$
2641bo33$3664b3o$3664bo1023b3o$3665bo1022bo$4689bo88$2768b3o$2768bo$
2769bo27$3792b3o$3792bo$3793bo8$4816b3o$4816bo$4817bo99$2896b3o$2896bo
1023b3o$2897bo1022bo$3921bo40$4944b3o$4944bo$4945bo75$3024b3o$3024bo$
3025bo29$4048b3o1021b3o$4048bo1023bo$4049bo1023bo83$3152b3o$3152bo$
3153bo31$4176b3o$4176bo$4177bo5$5200b3o$5200bo$5201bo112$3280b3o$3280b
o$3281bo14$4304b3o$4304bo$4305bo25$5328b3o$5328bo$5329bo81$3408b3o$
3408bo$3409bo19$4432b3o$4432bo$4433bo5$5456b3o$5456bo$5457bo96$3536b3o
$3536bo$3537bo26$4560b3o$4560bo$4561bo97$3664b3o$3664bo$3665bo23$4688b
3o$4688bo$4689bo93$3792b3o$3792bo$3793bo29$4816b3o$4816bo$4817bo79$
3920b3o$3920bo$3921bo38$4944b3o$4944bo$4945bo109$4048b3o$4048bo$4049bo
18$5072b3o$5072bo$5073bo107$4176b3o$4176bo$4177bo9$5200b3o$5200bo$
5201bo105$4304b3o$4304bo$4305bo28$5328b3o$5328bo$5329bo106$4432b3o$
4432bo$4433bo23$5456b3o$5456bo$5457bo92$4560b3o$4560bo$4561bo36$5584b
3o$5584bo$5585bo96$4688b3o$4688bo$4689bo23$5712b3o$5712bo$5713bo87$
4816b3o$4816bo$4817bo20$5840b3o$5840bo$5841bo101$4944b3o$4944bo$4945bo
135$5072b3o$5072bo$5073bo125$5200b3o$5200bo$5201bo136$5328b3o$5328bo$
5329bo109$5456b3o$5456bo$5457bo!
And here is the recipes array:

Code: Select all

[[-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -12, -2, -26, -12, -35, -44, -45, -27, -26, -11, -10, -27], [-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -21, -31, -18, -6, -12, 11, 7, 5, 3, 0, 8, -1, 0, 0, -10, -27], [-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -21, -30, -21, -25, -22, -31, -38, -41, -32, -18, -21, -30, -20, -10, -27], [-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -11, -10, -19, -28, -23, -24, -27, -22, -20, -30, -10, -10, -27], [-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -12, -2, -34, -4, -13, -22, -22, -20, -25, -13, -9, -10, -27], [-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -22, -23, -24, -24, -33, -33, -32, -16, -36, -39, -10, -27], [-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -12, -22, -13, -10, -5, -9, -11, -18, -15, -23, -14, -9, -5, -10, -27], [-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -11, -10, -20, -10, -19, -11, -25, -28, -19, -20, -10, -27]]
For completeness here is the recipe without adjustment but ends with block + hwss:

Code: Select all

[-4, -6, -8, -9, -17, -35, -37, -39, -47, -60, -55, -37, -35, -27, -41, -48, -41, -40, -30, -21, -21, -23, -11, -19, -31, -5, -7, -9, -10, -18, -34, -19, -22, -10, -27]
Last edited by simsim314 on June 24th, 2015, 3:18 am, edited 1 time in total.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 23rd, 2015, 4:32 pm

simsim314 wrote:
dvgrn wrote:eight equal-length timing adjustment tricks (whatever that means exactly -- I'm not quite clear on the details yet)
Well the details are very simple. Because the SLs can move only in even steps to fit the *WSS salvo "reader", so to fit the exact timing and parity of generated *WSS we need output glider to be "shifted" in phase and lane. You can look at input glider as the last glider in the slow-salvo recipe, and the output is what comes out from the one of 8 recipes.
I definitely understand the timing-adjustment idea. The way I think of it, to make the back end work right, at any given instant each *WSS in a downship stream has to be at a specific location relative to the front end. Same for upship streams relative to the back end.

If the caterloopillar is period N, there are N possible locations, and any given *WSS recipe will only have a 1/N chance of putting its *WSS in the correct location.

So we need timing-adjustment tricks to make up the difference. The vertical placement of the loaves is one variable -- though we may have to move a loaf a long way, and leave a lot of empty helix cycles, to get the right timing just by changing loaf positions. For some spaceship periods I'm guessing that trick won't work, though I haven't sorted out the math yet. Luckily glider-retimer recipes are another possible trick.

Maybe if the spaceship period is relatively prime to 8, we won't absolutely need glider-retimer recipes, if we don't mind the spaceship being a whole lot longer?
simsim314 wrote:I'm not sure if some trickery can be done by using odd number of SLs to make void operation, thus switching the parity of the input glider. I'm not sure if this trick will actually switch the lane, or maybe just do the same one lane lower.
You can definitely switch the lane -- or should we say, "change the lane parity"? -- with a void operation. That is, if you move a loaf from one helix cycle to the next, and leave the original helix cycle empty, you can access alternate glider lanes.

Here's a sample for the period 892 front end. Move the loaf down 236+N cells (with N odd) and you're suddenly firing gliders on lane N instead of lane 0. It's probably highly significant that 236 = 4*59...

Code: Select all

x = 413, y = 2660, rule = LifeHistory
6.C259.C$5.C.C257.C.C$5.C.C257.C.C$6.C259.C2$.2C7.2C249.2C7.2C$C2.C5.
C2.C247.C2.C5.C2.C$.2C7.2C249.2C7.2C2$6.C259.C$5.C.C257.C.C$5.C.C257.
C.C$6.C259.C5$9.3C257.3C$9.C2.C256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C
259.C$10.C.C257.C.C20$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.
3C$45.3C257.3C$45.3C257.3C$45.2C258.2C15$50.C259.C$49.3C257.3C$48.2C.
C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C7$200.F$200.
F$200.F$200.F$200.F$200.F$200.F$5.3C192.F64.3C$4.C2.C192.F63.C2.C$7.C
192.F66.C$7.C192.F66.C$7.C192.F66.C$6.C193.F65.C$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$55.C144.F114.C$54.3C
143.F113.3C$54.C.2C142.F113.C.2C$55.3C142.F114.3C$55.3C142.F114.3C$
55.2C143.F114.2C$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.
F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$63.C136.F122.C$62.3C135.F121.3C$62.C.2C134.F
121.C.2C$63.3C84.2A48.F122.3C84.2A$63.3C83.A2.A47.F122.3C83.A2.A$63.
2C85.2A48.F122.2C85.2A$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.
F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$71.C128.F130.C$70.3C127.F129.3C$70.C.2C
126.F129.C.2C$71.3C126.F130.3C$71.3C126.F130.3C$71.2C127.F130.2C$200.
F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F209.2A$200.F208.A2.A$59.3C138.F118.3C88.2A$59.
C2.C137.F118.C2.C$59.C140.F118.C$59.C3.C136.F118.C3.C$59.C3.C136.F
118.C3.C$59.C140.F118.C$60.C.C137.F119.C.C$200.F$200.F$200.F$200.F$
120.C79.F179.D$119.C.C78.F178.D.D$119.C2.C77.F178.D2.D$120.2C78.F179.
2D$200.F$200.F$200.F$79.C120.F138.C$78.3C119.F137.3C$78.C.2C118.F137.
C.2C$79.3C118.F138.3C$79.3C118.F138.3C$79.2C119.F138.2C$9.3C188.F68.
3C$9.C2.C187.F68.C2.C$9.C190.F68.C$9.C3.C186.F68.C3.C$9.C190.F68.C$
10.C.C187.F69.C.C$200.F$200.F$75.3C122.F134.3C$75.C2.C121.F134.C2.C$
75.C124.F134.C$75.C3.C120.F134.C3.C$75.C3.C120.F134.C3.C$75.C124.F
134.C$76.C.C121.F135.C.C$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$45.C154.F104.C$44.3C153.F103.3C$44.C.2C152.F103.C.
2C$45.3C152.F104.3C$45.3C152.F104.3C$45.3C152.F104.3C$45.2C153.F104.
2C$200.F$113.3C84.F172.3C34.2A$112.C2.C84.F171.C2.C33.A2.A$115.C84.F
174.C34.2A$111.C3.C84.F170.C3.C$115.C84.F174.C$112.C.C85.F171.C.C$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$50.C149.F109.C$49.3C148.F
108.3C$48.2C.C148.F107.2C.C$48.3C149.F107.3C$48.3C149.F107.3C$48.3C
149.F107.3C$49.2C149.F108.2C$200.F$200.F$200.F$200.F$200.F$200.F$200.
F$200.F$200.F$200.F$200.F$200.F$200.F$5.3C192.F64.3C$4.C2.C192.F63.C
2.C$7.C192.F66.C$7.C192.F66.C$7.C192.F66.C$6.C193.F65.C$200.F$200.F$
200.F$200.F$200.F$137.C62.F196.C$136.3C61.F195.3C$135.2C.C61.F194.2C.
C$135.3C62.F194.3C$117.3C15.3C62.F176.3C15.3C$117.C2.C14.3C62.F176.C
2.C14.3C$117.C18.2C62.F176.C18.2C$117.C3.C78.F176.C3.C$117.C3.C78.F
176.C3.C$117.C82.F176.C$118.C.C79.F177.C.C$200.F$200.F$200.F$200.F$
200.F209.2A$200.F208.A2.A$55.C144.F114.C94.2A$54.3C11.C131.F113.3C11.
C$54.C.2C9.3C130.F113.C.2C9.3C$55.3C9.C.2C129.F114.3C9.C.2C$55.3C10.
3C129.F114.3C10.3C$55.2C11.3C129.F114.2C11.3C$68.3C129.F127.3C$68.2C
130.F127.2C$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$84.C115.F
143.C$83.3C114.F142.3C$83.C.2C113.F142.C.2C$84.3C113.F143.3C$84.3C
113.F143.3C$84.3C113.F143.3C$84.2C114.F143.2C$200.F$98.C101.F157.C$
97.3C100.F156.3C$97.C.2C99.F156.C.2C$98.3C99.F157.3C$98.3C99.F157.3C$
98.3C99.F157.3C$63.C34.2C100.F122.C34.2C$62.3C135.F121.3C$62.C.2C134.
F121.C.2C$63.3C134.F122.3C$63.3C134.F122.3C$63.2C135.F122.2C$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F209.2A$
200.F208.A2.A$200.F209.2A$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$131.C68.F190.C$
130.3C67.F189.3C$130.C.2C66.F189.C.2C$131.3C66.F190.3C$131.3C66.F190.
3C$131.2C67.F190.2C$71.C128.F130.C$70.3C127.F129.3C$70.C.2C126.F129.C
.2C$71.3C46.C79.F130.3C46.C$71.3C45.3C78.F130.3C45.3C$71.2C46.C.2C77.
F130.2C46.C.2C$120.3C77.F179.3C$120.3C15.3C59.F179.3C15.3C$120.3C14.C
2.C59.F179.3C14.C2.C$120.2C18.C59.F179.2C18.C$136.C3.C59.F195.C3.C$
136.C3.C59.F195.C3.C$140.C59.F199.C$137.C.C60.F196.C.C$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$200.F$59.3C138.F118.3C$59.C2.C
137.F118.C2.C$59.C140.F118.C$59.C3.C136.F118.C3.C$59.C3.C136.F118.C3.
C$59.C140.F118.C$60.C.C137.F119.C.C$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$79.C120.F138.C$78.3C119.F137.3C$
78.C.2C118.F137.C.2C$79.3C118.F138.3C$79.3C118.F138.3C$79.2C119.F138.
2C$9.3C188.F68.3C$9.C2.C187.F68.C2.C$9.C190.F68.C$9.C3.C186.F68.C3.C
106.C$9.C190.F68.C109.C.C$10.C.C107.C79.F69.C.C106.C2.C$119.C.C78.F
179.2C$119.C2.C77.F$75.3C42.2C78.F134.3C$75.C2.C121.F134.C2.C$75.C
124.F134.C$75.C3.C120.F134.C3.C$75.C3.C120.F134.C3.C$75.C124.F134.C$
76.C.C121.F135.C.C$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
120.C79.F$119.C.C78.F$45.C73.C2.C77.F104.C$44.3C73.2C78.F103.3C$44.C.
2C152.F103.C.2C$45.3C152.F104.3C$45.3C152.F104.3C$45.3C152.F104.3C$
45.2C153.F104.2C$200.F$113.3C84.F172.3C$112.C2.C84.F171.C2.C$115.C84.
F174.C$111.C3.C84.F170.C3.C$115.C84.F174.C$112.C.C85.F171.C.C$200.F$
200.F$200.F$200.F$200.F$200.F$200.F$50.C149.F109.C$49.3C148.F108.3C$
48.2C.C148.F107.2C.C$48.3C149.F107.3C$48.3C149.F107.3C$48.3C149.F107.
3C$49.2C149.F108.2C$200.F$200.F$200.F$200.F$200.F$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$5.3C192.F64.3C$4.C2.C192.F63.C2.C$7.C
192.F66.C$7.C192.F66.C$7.C192.F66.C$6.C193.F65.C$200.F$200.F$200.F$
200.F$200.F$137.C62.F196.C$136.3C61.F195.3C$135.2C.C61.F194.2C.C$135.
3C62.F194.3C$117.3C15.3C62.F176.3C15.3C$117.C2.C14.3C62.F176.C2.C14.
3C$117.C18.2C62.F176.C18.2C$117.C3.C78.F176.C3.C$117.C3.C78.F176.C3.C
$117.C82.F176.C$118.C.C79.F177.C.C$200.F$200.F$200.F$200.F$200.F$200.
F$55.C144.F114.C$54.3C11.C131.F113.3C11.C$54.C.2C9.3C130.F113.C.2C9.
3C$55.3C9.C.2C129.F114.3C9.C.2C$55.3C10.3C129.F114.3C10.3C$55.2C11.3C
129.F114.2C11.3C$68.3C129.F127.3C$68.2C130.F127.2C$200.F$200.F$200.F$
200.F$200.F$200.F$200.F$200.F$200.F11$84.C259.C$83.3C257.3C$83.C.2C
256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.2C2$98.C259.C$
97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.3C257.3C$63.C
34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$63.3C257.3C$
63.2C258.2C28$145.2A$145.2A$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$119.D3C4.2A9.3C239.3C15.3C$119.D3C.2A11.C2.C239.
3C14.C2.C$120.2C2.A5.A9.C239.2C18.C$130.A5.C3.C255.C3.C$125.A4.A5.C3.
C255.C3.C$140.C259.C$126.3A8.C.C257.C.C10$59.3C257.3C$59.C2.C57.C198.
C2.C$59.C59.C.C197.C$59.C3.C55.C2.C196.C3.C$59.C3.C56.2C197.C3.C$59.C
259.C$60.C.C257.C.C12$79.C259.C$78.3C39.C217.3C$78.C.2C37.C.C216.C.2C
$79.3C37.C2.C216.3C$79.3C38.2C217.3C$79.2C258.2C$9.3C257.3C$9.C2.C
256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C259.C$10.C.C257.C.C3$75.3C257.3C$
75.C2.C256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$
76.C.C257.C.C11$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.3C$45.
3C257.3C$45.3C257.3C$45.2C258.2C2$113.3C257.3C$112.C2.C256.C2.C$115.C
259.C$111.C3.C255.C3.C$115.C259.C$112.C.C257.C.C8$50.C259.C$49.3C257.
3C$48.2C.C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C14$
5.3C257.3C$4.C2.C256.C2.C$7.C259.C$7.C259.C$7.C259.C$6.C259.C6$137.C
259.C$136.3C257.3C$135.2C.C256.2C.C$135.3C257.3C$117.3C15.3C239.3C15.
3C$117.C2.C14.3C239.C2.C14.3C$117.C18.2C239.C18.2C$117.C3.C255.C3.C$
117.C3.C255.C3.C$117.C259.C$118.C.C257.C.C7$55.C259.C$54.3C11.C245.3C
11.C$54.C.2C9.3C244.C.2C9.3C$55.3C9.C.2C244.3C9.C.2C$55.3C10.3C244.3C
10.3C$55.2C11.3C244.2C11.3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C
257.3C$83.C.2C256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.
2C2$98.C259.C$97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.
3C257.3C$63.C34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$
63.3C257.3C$63.2C258.2C30$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.
C$120.2C18.C239.2C18.C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$
137.C.C257.C.C10$59.3C257.3C$59.C2.C256.C2.C$59.C259.C$59.C3.C255.C3.
C$59.C3.C255.C3.C$59.C259.C$60.C.C257.C.C12$79.C259.C$78.3C257.3C$78.
C.2C256.C.2C$79.3C257.3C$79.3C257.3C$79.2C258.2C$9.3C257.3C$9.C2.C
256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C259.C$10.C.C257.C.C3$75.3C257.3C$
75.C2.C256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$
76.C.C257.C.C11$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.3C$45.
3C257.3C$45.3C257.3C$45.2C258.2C2$113.3C257.3C$112.C2.C256.C2.C$115.C
259.C$111.C3.C255.C3.C$115.C259.C$112.C.C257.C.C8$50.C259.C$49.3C257.
3C$48.2C.C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C14$
5.3C257.3C$4.C2.C256.C2.C$7.C259.C$7.C259.C$7.C259.C$6.C259.C6$137.C
259.C$136.3C257.3C$135.2C.C256.2C.C$135.3C257.3C$117.3C15.3C239.3C15.
3C$117.C2.C14.3C239.C2.C14.3C$117.C18.2C239.C18.2C$117.C3.C255.C3.C$
117.C3.C255.C3.C$117.C259.C$118.C.C257.C.C7$55.C259.C$54.3C11.C245.3C
11.C$54.C.2C9.3C244.C.2C9.3C$55.3C9.C.2C244.3C9.C.2C$55.3C10.3C244.3C
10.3C$55.2C11.3C244.2C11.3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C
257.3C$83.C.2C256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.
2C2$98.C259.C$97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.
3C257.3C$63.C34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$
63.3C257.3C$63.2C258.2C30$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.
C$120.2C18.C239.2C18.C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$
137.C.C257.C.C10$59.3C257.3C$59.C2.C256.C2.C$59.C259.C$59.C3.C255.C3.
C$59.C3.C255.C3.C$59.C259.C$60.C.C257.C.C12$79.C259.C$78.3C257.3C$78.
C.2C256.C.2C$79.3C257.3C$79.3C257.3C$79.2C258.2C$9.3C257.3C$9.C2.C
256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C259.C$10.C.C257.C.C3$75.3C257.3C$
75.C2.C256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$
76.C.C257.C.C11$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.3C$45.
3C257.3C$45.3C257.3C$45.2C258.2C2$113.3C257.3C$112.C2.C256.C2.C$115.C
259.C$111.C3.C255.C3.C$115.C259.C$112.C.C257.C.C8$50.C259.C$49.3C257.
3C$48.2C.C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C14$
5.3C257.3C$4.C2.C256.C2.C$7.C259.C$7.C259.C$7.C259.C$6.C259.C6$137.C
259.C$136.3C257.3C$135.2C.C256.2C.C$135.3C257.3C$117.3C15.3C239.3C15.
3C$117.C2.C14.3C239.C2.C14.3C$117.C18.2C239.C18.2C$117.C3.C255.C3.C$
117.C3.C255.C3.C$117.C259.C$118.C.C257.C.C7$55.C259.C$54.3C11.C245.3C
11.C$54.C.2C9.3C244.C.2C9.3C$55.3C9.C.2C244.3C9.C.2C$55.3C10.3C244.3C
10.3C$55.2C11.3C244.2C11.3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C
257.3C$83.C.2C256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.
2C2$98.C259.C$97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.
3C257.3C$63.C34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$
63.3C257.3C$63.2C258.2C30$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.
C$120.2C18.C239.2C18.C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$
137.C.C257.C.C10$59.3C257.3C$59.C2.C256.C2.C$59.C259.C$59.C3.C255.C3.
C$59.C3.C255.C3.C$59.C259.C$60.C.C257.C.C12$79.C259.C$78.3C257.3C$78.
C.2C256.C.2C$79.3C257.3C$79.3C257.3C$79.2C258.2C$9.3C257.3C$9.C2.C
256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C259.C$10.C.C257.C.C3$75.3C257.3C$
75.C2.C256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$
76.C.C257.C.C11$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.3C$45.
3C257.3C$45.3C257.3C$45.2C258.2C2$113.3C257.3C$112.C2.C256.C2.C$115.C
259.C$111.C3.C255.C3.C$115.C259.C$112.C.C257.C.C8$50.C259.C$49.3C257.
3C$48.2C.C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C14$
5.3C257.3C$4.C2.C256.C2.C$7.C259.C$7.C259.C$7.C259.C$6.C259.C6$137.C
259.C$136.3C257.3C$135.2C.C256.2C.C$135.3C257.3C$117.3C15.3C239.3C15.
3C$117.C2.C14.3C239.C2.C14.3C$117.C18.2C239.C18.2C$117.C3.C255.C3.C$
117.C3.C255.C3.C$117.C259.C$118.C.C257.C.C7$55.C259.C$54.3C11.C245.3C
11.C$54.C.2C9.3C244.C.2C9.3C$55.3C9.C.2C244.3C9.C.2C$55.3C10.3C244.3C
10.3C$55.2C11.3C244.2C11.3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C
257.3C$83.C.2C256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.
2C2$98.C259.C$97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.
3C257.3C$63.C34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$
63.3C257.3C$63.2C258.2C30$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.
C$120.2C18.C239.2C18.C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$
137.C.C257.C.C10$59.3C257.3C$59.C2.C256.C2.C$59.C259.C$59.C3.C255.C3.
C$59.C3.C255.C3.C$59.C259.C$60.C.C257.C.C12$79.C259.C$78.3C257.3C$78.
C.2C256.C.2C$79.3C257.3C$79.3C257.3C$79.2C258.2C$9.3C257.3C$9.C2.C
256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C259.C$10.C.C257.C.C3$75.3C257.3C$
75.C2.C256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$
76.C.C257.C.C11$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.3C$45.
3C257.3C$45.3C257.3C$45.2C258.2C2$113.3C257.3C$112.C2.C256.C2.C$115.C
259.C$111.C3.C255.C3.C$115.C259.C$112.C.C257.C.C8$50.C259.C$49.3C257.
3C$48.2C.C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C14$
5.3C257.3C$4.C2.C256.C2.C$7.C259.C$7.C259.C$7.C259.C$6.C259.C6$137.C
259.C$136.3C257.3C$135.2C.C256.2C.C$135.3C257.3C$117.3C15.3C239.3C15.
3C$117.C2.C14.3C239.C2.C14.3C$117.C18.2C239.C18.2C$117.C3.C255.C3.C$
117.C3.C255.C3.C$117.C259.C$118.C.C257.C.C7$55.C259.C$54.3C11.C245.3C
11.C$54.C.2C9.3C244.C.2C9.3C$55.3C9.C.2C244.3C9.C.2C$55.3C10.3C244.3C
10.3C$55.2C11.3C244.2C11.3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C
257.3C$83.C.2C256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.
2C2$98.C259.C$97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.
3C257.3C$63.C34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$
63.3C257.3C$63.2C258.2C30$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.
C$120.2C18.C239.2C18.C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$
137.C.C257.C.C10$59.3C257.3C$59.C2.C256.C2.C$59.C259.C$59.C3.C255.C3.
C$59.C3.C255.C3.C$59.C259.C$60.C.C257.C.C12$79.C259.C$78.3C257.3C$78.
C.2C256.C.2C$79.3C257.3C$79.3C257.3C$79.2C258.2C$9.3C257.3C$9.C2.C
256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C259.C$10.C.C257.C.C3$75.3C257.3C$
75.C2.C256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$
76.C.C257.C.C11$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.3C$45.
3C257.3C$45.3C257.3C$45.2C258.2C2$113.3C257.3C$112.C2.C256.C2.C$115.C
259.C$111.C3.C255.C3.C$115.C259.C$112.C.C257.C.C8$50.C259.C$49.3C257.
3C$48.2C.C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C14$
5.3C257.3C$4.C2.C256.C2.C$7.C259.C$7.C259.C$7.C259.C$6.C259.C6$137.C
259.C$136.3C257.3C$135.2C.C256.2C.C$135.3C257.3C$117.3C15.3C239.3C15.
3C$117.C2.C14.3C239.C2.C14.3C$117.C18.2C239.C18.2C$117.C3.C255.C3.C$
117.C3.C255.C3.C$117.C259.C$118.C.C257.C.C7$55.C259.C$54.3C11.C245.3C
11.C$54.C.2C9.3C244.C.2C9.3C$55.3C9.C.2C244.3C9.C.2C$55.3C10.3C244.3C
10.3C$55.2C11.3C244.2C11.3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C
257.3C$83.C.2C256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.
2C2$98.C259.C$97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.
3C257.3C$63.C34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$
63.3C257.3C$63.2C258.2C30$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.
C$120.2C18.C239.2C18.C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$
137.C.C257.C.C10$59.3C257.3C$59.C2.C256.C2.C$59.C259.C$59.C3.C255.C3.
C$59.C3.C255.C3.C$59.C259.C$60.C.C257.C.C12$79.C259.C$78.3C257.3C$78.
C.2C256.C.2C$79.3C257.3C$79.3C257.3C$79.2C258.2C$9.3C257.3C$9.C2.C
256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C259.C$10.C.C257.C.C3$75.3C257.3C$
75.C2.C256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$
76.C.C257.C.C11$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.3C$45.
3C257.3C$45.3C257.3C$45.2C258.2C2$113.3C257.3C$112.C2.C256.C2.C$115.C
259.C$111.C3.C255.C3.C$115.C259.C$112.C.C257.C.C8$50.C259.C$49.3C257.
3C$48.2C.C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C14$
5.3C257.3C$4.C2.C256.C2.C$7.C259.C$7.C259.C$7.C259.C$6.C259.C6$137.C
259.C$136.3C257.3C$135.2C.C256.2C.C$135.3C257.3C$117.3C15.3C239.3C15.
3C$117.C2.C14.3C239.C2.C14.3C$117.C18.2C239.C18.2C$117.C3.C255.C3.C$
117.C3.C255.C3.C$117.C259.C$118.C.C257.C.C7$55.C259.C$54.3C11.C245.3C
11.C$54.C.2C9.3C244.C.2C9.3C$55.3C9.C.2C244.3C9.C.2C$55.3C10.3C244.3C
10.3C$55.2C11.3C244.2C11.3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C
257.3C$83.C.2C256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.
2C2$98.C259.C$97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.
3C257.3C$63.C34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$
63.3C257.3C$63.2C258.2C30$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.
C$120.2C18.C239.2C18.C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$
137.C.C257.C.C10$59.3C257.3C$59.C2.C256.C2.C$59.C259.C$59.C3.C255.C3.
C$59.C3.C255.C3.C$59.C259.C$60.C.C257.C.C12$79.C259.C$78.3C257.3C$78.
C.2C256.C.2C$79.3C257.3C$79.3C257.3C$79.2C258.2C$9.3C257.3C$9.C2.C
256.C2.C$9.C259.C$9.C3.C255.C3.C$9.C259.C$10.C.C257.C.C3$75.3C257.3C$
75.C2.C256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$
76.C.C257.C.C11$45.C259.C$44.3C257.3C$44.C.2C256.C.2C$45.3C257.3C$45.
3C257.3C$45.3C257.3C$45.2C258.2C2$113.3C257.3C$112.C2.C256.C2.C$115.C
259.C$111.C3.C255.C3.C$115.C259.C$112.C.C257.C.C8$50.C259.C$49.3C257.
3C$48.2C.C256.2C.C$48.3C257.3C$48.3C257.3C$48.3C257.3C$49.2C258.2C14$
5.3C257.3C$4.C2.C256.C2.C$7.C259.C$7.C259.C$7.C259.C$6.C259.C6$137.C
259.C$136.3C257.3C$135.2C.C256.2C.C$135.3C257.3C$117.3C15.3C239.3C15.
3C$117.C2.C14.3C239.C2.C14.3C$117.C18.2C239.C18.2C$117.C3.C255.C3.C$
117.C3.C255.C3.C$117.C259.C$118.C.C257.C.C7$55.C259.C$54.3C11.C245.3C
11.C$54.C.2C9.3C244.C.2C9.3C$55.3C9.C.2C244.3C9.C.2C$55.3C10.3C244.3C
10.3C$55.2C11.3C244.2C11.3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C
257.3C$83.C.2C256.C.2C$84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.
2C2$98.C259.C$97.3C257.3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.
3C257.3C$63.C34.2C223.C34.2C$62.3C257.3C$62.C.2C256.C.2C$63.3C257.3C$
63.3C257.3C$63.2C258.2C30$131.C259.C$130.3C257.3C$130.C.2C256.C.2C$
131.3C257.3C$131.3C257.3C$131.2C258.2C$71.C259.C$70.3C257.3C$70.C.2C
256.C.2C$71.3C46.C210.3C46.C$71.3C45.3C209.3C45.3C$71.2C46.C.2C208.2C
46.C.2C$120.3C257.3C$120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.
C$120.2C18.C239.2C18.C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$
137.C.C257.C.C10$59.3C257.3C$59.C2.C256.C2.C$59.C259.C$59.C3.C255.C3.
C$59.C3.C255.C3.C$59.C259.C$60.C.C257.C.C12$79.C259.C$78.3C257.3C$78.
C.2C256.C.2C$79.3C257.3C$79.3C257.3C$79.2C258.2C9$75.3C257.3C$75.C2.C
256.C2.C$75.C259.C$75.C3.C255.C3.C$75.C3.C255.C3.C$75.C259.C$76.C.C
257.C.C19$113.3C257.3C$112.C2.C256.C2.C$115.C259.C$111.C3.C255.C3.C$
115.C259.C$112.C.C257.C.C39$137.C259.C$136.3C257.3C$135.2C.C256.2C.C$
135.3C257.3C$117.3C15.3C239.3C15.3C$117.C2.C14.3C239.C2.C14.3C$117.C
18.2C239.C18.2C$117.C3.C255.C3.C$117.C3.C255.C3.C$117.C259.C$118.C.C
257.C.C8$68.C259.C$67.3C257.3C$67.C.2C256.C.2C$68.3C257.3C$68.3C257.
3C$68.3C257.3C$68.2C258.2C20$84.C259.C$83.3C257.3C$83.C.2C256.C.2C$
84.3C257.3C$84.3C257.3C$84.3C257.3C$84.2C258.2C2$98.C259.C$97.3C257.
3C$97.C.2C256.C.2C$98.3C257.3C$98.3C257.3C$98.3C257.3C$98.2C258.2C35$
131.C259.C$130.3C257.3C$130.C.2C256.C.2C$131.3C257.3C$131.3C257.3C$
131.2C258.2C4$120.C259.C$119.3C257.3C$119.C.2C256.C.2C$120.3C257.3C$
120.3C15.3C239.3C15.3C$120.3C14.C2.C239.3C14.C2.C$120.2C18.C239.2C18.
C$136.C3.C255.C3.C$136.C3.C255.C3.C$140.C259.C$137.C.C257.C.C!
Lower down on the left is an experiment with putting loaves closer together. It looks as if 18 cells vertically is the minimum spacing, so for some recipes it might be possible to pack things in very closely, with more than one loaf per helix cycle... it would be a little weird, though, since the lower-lane gliders start hitting the target area before the upper-lane gliders do.
simsim314 wrote:Meanwhile I see that there is no point in switching the parity, not sure exactly if this universally correct, or just for c/11 caterloopillar.
If you can switch the parity reliably by doing something short and simple, then it seems to me that that will cut in half the number of glider emitter variations that you'll need for that particular period of spaceship. For period 8N spaceships, the full eight emitters will pretty definitely be needed, though.

I could be horribly wrong about most of this -- but I'm looking forward to understanding it all eventually!

Currently I'm thinking that a mod-8 glider emitter adjustment toolkit will need to have a fixed height, but not a fixed number of gliders. Just pick some reasonable number of helix cycles, so that there's enough room fit any of the eight recipes that will be needed. Each N-mod-8 emitter will require an input column of blocks on a specific lane -- can be anywhere on the diagonal, but always the same lane.

All emitters produce a single output glider on some particular lane (always the same lane, relative to the blocks) but each emitter produces a different timing mod 8. The blocks are used up, and any leftover junk is cleaned up.

A toolkit like this would make swapping emitters relatively trivial. But maybe that's too expensive, and there's an easier way to get all the timing right...?

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 23rd, 2015, 4:49 pm

dvgrn wrote:Here's a sample for the period 892 front end...
... I believe I meant "the period 564 front end" -- i.e., the one posted here originally, and I think the slow-salvo generator script also uses the same period.

I got myself confused for a while, because helix cycle N+1 doesn't overlap with helix cycle N (at T=446, the *WSSes are all in the same location but opposite phases.) So I was counting the ticks until helix cycle N+2 overlaps with the original position of helix cycle N -- which is kind of irrelevant.

The overlap mismatch is exactly what I should be expecting if each new upship or downship is being created 59 cells higher than the last one -- but I still managed to be surprised for a few minutes there.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 23rd, 2015, 6:10 pm

simsim314 wrote:Here are recipes for all 8 adjustments + hwss (we need only 7, but I did the one that kept the state and parity by mistake so I post it as well)...
I don't quite trust these recipes yet, to really produce eight different mod-8 timings for all possible caterloopillar periods.

In the first recipe, for example, glider #40 is the trigger glider that actually produces the downward HWSS. In the second recipe, it's glider #43 instead. Unless the spaceship period is a multiple of 8, it seems to me that different gliders will produce different output timings.

So it will probably be possible to pick a spaceship period for which (Recipe #1 glider #40) produces an HWSS with the same exact timing as (Recipe #2 glider #43). Which would mean that some other timing adjustment won't be possible with this toolkit.

-- At least according to my crazy theories. Is my Visualization of the Cosmic All failing me again?

Also, the first and second recipes seem to produce HWSSes that are offset from each other by 48 cells, which is an even number. Now, we can throw in an empty helix cycle and get +/-59 spaces vertically (right? maybe?) so really that's the same as an offset of minus-11, which is a different phase after all... but how could alternative HWSSes be made that are offset by just 1 cell vertically?

I could imagine that it would be simpler to have recipes where the HWSS is always triggered by the same glider, Glider #T, and the resulting HWSSes are in adjacent spacetime locations -- 0, 1, 2, ... 7 mod 8. But that's likely to be a much more expensive toolkit, I suppose!

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 23rd, 2015, 6:54 pm

Well for now I just want to build c/11. If I get the timing tricks there, I will switch to even periods as they all have their own nuances. As I mentioned earlier for the odd period, the number of gliders didn't matter (or at lest they didn't yield new relative positions between two *WSS).

Replacing the recipes I posted with recipes where the block move take extra glider is simple adjustment (just move some deleted block extra iteration). On the other hand having odd period means that each next glider will be emitted in new internal state (I think), so maybe we need to have all recipes in length of mod 4.

I really need to figure out those details step by step, and hopefully will get the entire picture at the end. There are always many small nuances I certainly can't think about before I do some "testings" on specific cases.

EDIT Those recipes are certainly have the advantage of having all the 8 options relative to the trigger. If some external considerations require more gliders or some SL movement tricks, this could be done "on top" of these recipes.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 24th, 2015, 4:53 pm

dvgrn, several days ago, wrote:
simsim314 wrote:And in what order the slow salvo should work.
I find it a little hard to even consider any *WSS-stream construction order other than farthest to nearest.
...
The targets will have to get pushed all the way across all of the downship-stream columns before construction starts -- right?
No, my June 19th post was wrong, actually -- EDIT: or rather, half wrong. The above is correct for the downship stream construction, but wrong for the upship streams.

The construction is farthest-to-nearest in the sense that the most distant *WSS streams, both upships and downships, will already be present nearby while the nearer *WSS streams are being constructed. Scanning down the Caterloopillar from the top, the far rightmost downship stream will be created "first", by gliders hitting a target still life pushed far to the right by the first few dozen glider streams from the positive helix. Leftover debris from that first construction is then pulled incrementally inward and used to construct the remaining downship streams, one after the other.

However, for the upship streams, the most distant downship stream is built using leftover targets from the construction of the next nearer downship stream, which is constructed from debris from the next nearer one, and so on.

Here's a rough diagram. (I've been needing one of these for a while, so hopefully it's fairly close to correct and other people will also find it useful.)

EDIT: Fixed target push at the top of the image, based on simsim314's notes in the next post:
Rough blueprint for strange-loop Caterpillar, version 1.1
Rough blueprint for strange-loop Caterpillar, version 1.1
Caterloopillar.png (55.79 KiB) Viewed 19150 times
Each colored group of glider streams (dotted lines) builds one upship or downship stream. The circles are the locations of intermediate target constellations. The slow-salvo recipes gradually push the targets away as they're building successive *WSS streams.

The loaf still lifes that are read by each helix to produce the glider streams are marked in red, but really one filled red symbol corresponds to a long series of loaves. In the actual caterloopillar there will be a lot more gliders per *WSS, not just four, and over twice as many *WSS streams in all.

Also, there will really not be any big gap between the left and right sides, so the initial purple targets will be generated by the right half and the initial pinkish targets will be produced by the left half (the extra "edgy" SLs).

So... there's a very important sense in which the slow-salvo construction for downship streams is nearest-to-farthest. Ultimately there will be a slow-salvo recipe aimed at a single block that converts it into 16 successive downships. The upship half of the caterloopillar (the positive helix) will shoot that single recipe over and over again -- and that recipe will build the nearest *WSS first and the farthest one last. You have to scan the positive helix from the bottom up to see the targets and downship outputs in chronological order.

This is confusingly backward-and-yet-the-same from the negative-helix, downship half of the caterloopillar. Its slow salvo recipe will also build its 19 upship streams in nearest-to-farthest order, so here there's no need for a long push for an initial target. This time you can scan the glider streams and targets from the top down to see the slow salvo in the correct chronological order.

[Okay, how much did I get wrong?]

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 24th, 2015, 6:05 pm

dvgrn wrote:[Okay, how much did I get wrong?]
I think half of it :)

First of all the picture is almost correct.

For the down stream you right we need to push it as you described. But the up stream starts at the head very close to the cateloopillar spine. If you will look what is closest to the head at your sketch, is actually the one which is furthest from the spine. So the up stream will need to push the block all the way to the other side unlike the down stream. So to fix your picture you need to add the push at the front.

The difference between the up and down stream is very confusing. The point is that slow salvo always built from top to bottom, the reason for this is because the caterloopillar moves up, so the new appeared SL will always be upper to the previous one. But for the down stream that means exactly the opposite than what it means to the up stream. So all the actions and the construction logic is exactly opposite.

EDIT There new SL should be created very fast, so there is no time to wait until some recipe from the bottom will push it to the new lane for the upper stream. One should create the new SL from the head.
Attachments
Caterloopillar.png
Caterloopillar.png (108.46 KiB) Viewed 19163 times

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 24th, 2015, 8:44 pm

simsim314 wrote:
dvgrn wrote:[Okay, how much did I get wrong?]
I think half of it :)

First of all the picture is almost correct.
Ah, that's not too bad, then... I've added a target-push section at the top of the image, and edited the explanations accordingly -- thanks for the clarification!

I'm really enjoying this impressive strange-loop design. It can be added to the list as the twentieth Life construction project that uses slow-salvo technology. So far no two projects have been able to use quite the same toolkit; there's something significantly different about each one.

-- What's the minimum distance between cycles for the two step-59 helices, by the way? c/11 looks like the fastest odd c/N period that can be supported, but that's all I've figured out so far.

EDIT: If the leftover junk constellations from the last downship stream synthesis on the right don't get cleaned up, could they be used as targets by the negative-helix slow salvo? (I.e.,. make all the target dots into a connected line.) I don't think it would save any *WSSes, but it would be kind of elegant -- in an awkward slow-salvo-ish way.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 25th, 2015, 3:09 am

dvgrn wrote:I've added a target-push section at the top of the image
Everything looks correct now.
dvgrn wrote:I'm really enjoying this impressive strange-loop design.
Well the idea is pretty obvious, but I should mention Alexey Nigin for bringing it up and clearly formulating it. I was contemplating it from time to time, but it seemed impossible to me for some (wrong) reason at the time.
See this and this. I also made quiet few calculations to make sure this is the cheapest design, and it's definitely the most elegant.
dvgrn wrote: It can be added to the list as the twentieth Life construction project that uses slow-salvo technology
Your post made me think. If we want to have any speed above c/4 we need to start devising monochromatic slow salvo recipes. The movement of SL by even step should be pretty simple, but this will mean we need good collection of monochromatic slow salvo recipes.

There is no way to avoid even movement. Lets say I want to have speed 40/401, so the block needs to move 40 * N and the period would be 401 * N, but 40 * N is always even meaning I must use monochromatic slow salvo otherwise I'll only afford (2k + 1) / N speeds only.

EDIT Thinking a bit more, we need only monochromatic slow salvo. If we want to have speed of (2k + 1) / N we can just make it (4k + 2) / 2N. So even steps will include odd speeds but odd will not include even.
dvgrn wrote:make all the target dots into a connected line.
I can't see why not. I think for my initial sketches I'll start without this trick (just to concentrate on the major issues, and I already have all the recipes necessary for complete separation between up and down streams), but this looks to me as the first simple enough improvement which will save HWSS from the downstream.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 25th, 2015, 7:09 am

simsim314 wrote:If we want to have any speed above c/4 ... [t]here is no way to avoid even movement. Lets say I want to have speed 40/401, so the block needs to move 40 * N and the period would be 401 * N, but 40 * N is always even meaning I must use monochromatic slow salvo otherwise I'll only afford (2k + 1) / N speeds only.
Yes, I guess if you're aiming for a hyperadjustable spaceship-building toolkit that can attain any sufficiently slow rational speed, P1 monochromatic is the only way to go. Silly me, I thought you might stop with "any c/N speed, N>9", or something like that (since you can hit that series easily with 59c/59N caterloopillars).

Luckily, P1 monochromatic slow salvos are provably universal. At worst, you can build freeze-dried slow salvos out of P1 junk. Those can then build and trigger any *WSS seed recipe you want, including the more efficient P2 polychromatic recipes from the Centipede.

But that's just an upper bound. Probably a direct search will turn up much smaller *WSS edge-shooter and timing adjustment recipes.
simsim314 wrote:EDIT Thinking a bit more, we need only monochromatic slow salvo. If we want to have speed of (2k + 1) / N we can just make it (4k + 2) / 2N. So even steps will include odd speeds but odd will not include even.
I take it this might mean going back to the old 38-step helices at some point. But it looks like that's a fixed recipe -- can't be adjusted trivially to make 38+4N or whatever. Do you have a small adjustable even-step helix?

Also, might it still be worth looking for more efficient cleanup tricks, for the 38-step helix or the 59-step one? Or do the 59-step helices-with-cleanup adjust easily to 59+X, and it's not worth looking for over-specific tricks that would (probably) break that? I haven't quite dared to try moving all those streams around yet...!

EDIT: Now that I look again, the 59-step negative helix doesn't look particularly easy to adjust to a different odd step size (?) so maybe that's not part of the plan anyway.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 25th, 2015, 1:41 pm

dvgrn wrote:Silly me, I thought you might stop with "any c/N speed, N>9"
Hehe why stop there if just another step will bring us to all the speeds there are bellow c/4.
dvgrn wrote: P1 monochromatic slow salvos are provably universal. At worst, you can build freeze-dried slow salvos out of P1 junk.
It's not the question of whether it's possible, but how much time will take to write the script that does it, and run it. I guess my Glue script can be modified to use monochromatic salvos. I will post it here as soon as I will get around it.
dvgrn wrote:I take it this might mean going back to the old 38-step helices at some point.
If anything I'll probably prefer the 58 step here
dvgrn wrote:But it looks like that's a fixed recipe -- can't be adjusted trivially to make 38+4N or whatever.
All of it is pretty trivial. Converting SL->glider and glider to SL in all parities is pretty simple, and I remember there are a lot of options there. I think in Jason's collection there are all parities *WSS salvo -> glider reflectors already.
dvgrn wrote:Do you have a small adjustable even-step helix?
The helices I used in the head an tail recipes are trivially adjustable to even step. There are just two ways (one odd and one even) to make BH from glider+*WSS collision, the same goes for HF. This is why it was so simple for me to find the recipes in the first place - they work in both parities.
dvgrn wrote:Also, might it still be worth looking for more efficient cleanup tricks
Definitely I can reduce the cleanup by factor of 2. One just need to place the *WSS colliding with the glider emitted from the helix in front of the stream instead from the side. There is glider + *WSS -> block + glider recipe, that can be used for cleaning.

Few examples from glider+MWSS collision search (HWSS has more):

Code: Select all

x = 74, y = 19, rule = LifeHistory
11.3C57.3C$11.C59.C$12.C59.C$61.3C$61.C2.C$61.C$61.C3.C$61.C$62.C.C5$
2.C$.3C$2C.C$3C$3C$.2C!
The only reason I'm using not the "in-front" approach, is because this will force the distance between the two reading heads larger and will probably forbid some high speeds at the 59 case. Switching to more general case I'll definitely redesign the deletion approach.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 25th, 2015, 5:24 pm

Here's an interesting item that showed up in a five-minute search with a modified version of chris_c's script. Will this be too slow of an MWSS edge-shooting reaction for any reason?

Code: Select all

x = 176, y = 160, rule = LifeHistory
173.2A$173.2A3$173.3A$175.A$174.A7$156.3A$158.A$157.A13$157.3A$159.A$
158.A4$146.3A$148.A$131.3A13.A$133.A$132.A7$120.3A$122.A$121.A81$21.
3A$23.A$22.A14$10.3A$12.A$11.A11$3A$2.A$.A!
It looks as if it should work okay -- and in any case I think that settling too slowly is not a problem for this project, as long as the active reaction stays out of the way of the *WSS lane. If a longer gap is needed, moving the SL down a couple of helix cycles should give plenty of extra time.

Unfortunately two of the gliders are different colors from the rest, so it would be fairly expensive to retool this into a monochromatic P1 recipe. The block-pull glider could probably be replaced, with maybe a dozen monochromatic gliders... but the second-to-last glider is part of the Blinker Cleanup Team, and it looks like it really needs to be on that lane.

-- I know blinkers are ordinarily forbidden in this project, but this one piece of leftover junk is actually okay. Any glider in the recipe can be any phase, and the cleanup will still work:

Code: Select all

x = 347, y = 185, rule = LifeHistory
172.2B158.2B$171.4B156.4B$171.4B156.4B$170.6B154.6B$169.4B2DB153.7B$
168.5B2DB152.8B$167.9B151.9B$167.10B4.B145.9B5.B$166.6B2D3B3.3B.2B
140.9BD4.3B.2B$166.6B2D13B139.8BDBD10B$165.22B138.8BD2BD10B$162.25B
135.12B2D11B$161.26B134.26B$161.26B134.26B$160.27B133.27B$161.25B135.
25B10$173.2A158.2A$173.2A158.2A3$173.2A158.3A$174.2A159.A$173.A160.A
7$156.2A158.3A$157.2A159.A$156.A160.A13$157.2A158.3A$158.2A159.A$157.
A160.A4$146.2A158.3A$147.2A159.A$131.2A13.A144.3A13.A$132.2A159.A$
131.A160.A7$120.3A157.3A$122.A159.A$121.A159.A81$21.2A158.3A$22.2A
159.A$21.A160.A14$10.2A158.3A$11.2A159.A$10.A160.A11$2A158.3A$.2A159.
A$A160.A!
#C [[ STOP 660 ]]
EDIT:There's another option, but on another wrong-parity lane:

Code: Select all

x = 336, y = 160, rule = LifeHistory
173.2A158.2A$173.2A158.2A3$173.2A158.3A$174.2A159.A$173.A160.A7$156.
2A158.3A$157.2A159.A$156.A160.A13$157.2A158.3A$158.2A159.A$157.A160.A
4$146.2A158.3A$147.2A159.A$131.2A13.A144.3A13.A$132.2A159.A$131.A160.
A7$120.3A157.3A$122.A159.A$121.A159.A81$21.2A158.3A$22.2A159.A$21.A
160.A14$6.2A158.3A$7.2A159.A$6.A160.A11$2A158.3A$.2A159.A$A160.A!
This nine-glider recipe was a lucky adaptation of a five-glider P2-slow version that showed up on a test run of the search script:

Code: Select all

x = 85, y = 73, rule = B3/S23
63b2o$63b2o6$51b3o$53bo$52bo30b2o$83b2o13$36b3o$38bo$37bo17$14bo$14b2o
$13bobo15$9b3o$11bo$10bo9$3o$2bo$bo!
I also thought of pre-placing a block to get rid of the blinker, as shown, but luckily there turned out to be an easy period-independent cleanup.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 25th, 2015, 5:48 pm

dvgrn wrote:Here's an interesting item that showed up in a five-minute search with a modified version of chris_c's script. Will this be too slow of an MWSS edge-shooting reaction for any reason?
Wow, great! This really reduces MWSS a lot. Can you please post the lane array as well?

Anyway I've committed to GitHub my monochromatic Glue script, with result up to depth 10 (the branching factor was 3.5 so depth 10 is not that much). Check it out.

If you say it took only 5 minutes to find the MWSS edge shooter, makes me really consider to try and adopt my script for edge shooters as well.

EDIT I really liked the trick with the blinker, kinda of unexpected. In general we can use Glue to track both states of a blinker, and find recipe that will delete the blinker and the TL. Obviously some more complex constellations of blinker + SL or partial TL etc. could appear but in this case we could run the script once more for that constellation.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 25th, 2015, 8:21 pm

simsim314 wrote:Wow, great! This really reduces MWSS a lot. Can you please post the lane array as well?
Well, naturally chris_c's script uses a different numbering system from your Glue script -- -5 and 4 are the honeyfarm lanes instead of your -4 and 5. I like Paul Chapman's lane numbering even more now. I believe the recipe in your notation would be

5, -3, 13, 8, -5, -7, NOP, -23, -18, -15

(the NOP is maybe not needed -- it just marks the spot where a delay of several hundred ticks is required.)
simsim314 wrote:Anyway I've committed to GitHub my monochromatic Glue script, with result up to depth 10 (the branching factor was 3.5 so depth 10 is not that much).
Looks good! Any plans for a two-week run this time around, to get a wider variety of recipes for some of the less common still lifes?
simsim314 wrote:If you say it took only 5 minutes to find the MWSS edge shooter, makes me really consider to try and adopt my script for edge shooters as well.
It was just a few minutes with a not particularly optimized Python script, yes -- but that's only because there happened to be a five-glider edge shooter out there in the bigger search space. The leftover junk is 28x27. And it was really really lucky that the P2 TL->block placement could be replaced with a simple beehive transformation and block pull... but a P1 search should find the 6-glider version pretty quick, anyway.

There were actually four LWSS recipes that came before that, just four gliders away from a block -- one output in each direction (but none of them edge shooters). Just two recipes and their mirror images, really, but it makes a nice test of my *WSS detection code.
simsim314 wrote:EDIT I really liked the trick with the blinker, kinda of unexpected. In general we can use Glue to track both states of a blinker, and find recipe that will delete the blinker and the TL. Obviously some more complex constellations of blinker + SL or partial TL etc. could appear but in this case we could run the script once more for that constellation.
I think you mean "delete the blinker and the SL"?

Well, you can't just run the cleanup script again for a new P2 derivative constellation. With P2 debris changing to P2 debris you'd have to come up with a salvo that handles more different cases (incoming gliders with same parity vs. opposite parities, and even or odd parity first).

No doubt there are some really impressive Schroedinger's Recipes out there, that get to the same clean slate by several amusingly different routes -- but it probably won't usually be worth looking for them.

Two variations that are maybe worth looking for are the "detonators" (see this previous post) and the "backstops": if there's a still life directly or partly behind a blinker, even a long distance away, then it's often possible to [kill|modify] the blinker with the first glider, [modify the SL|clean up the blinker] with the second glider, and clean up the [unaltered|modified] SL with the third glider.

That's about as many superimposed quantum states as I want to think about at the moment, but I'm sure variants can be found that use more gliders, with just a little searching...!

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 26th, 2015, 5:38 am

dvgrn wrote: I believe the recipe in your notation would be
Thx, it worked.
dvgrn wrote: With P2 debris changing to P2 debris you'd have to come up with a salvo that handles more different cases (incoming gliders with same parity vs. opposite parities, and even or odd parity first).
Well if so we can limit the number of parallel options, in this "Schrodinger glue" (if there are too many you don't continue to evolve it), it would be interesting to see whether there is a full "collapse", in all possible input parities (I think that blinker and traffic light are the most natural candidates to start with).

One can use some tricks in our case, for example the difference between even and odd period is that odd periods are alternating. But if you make sure all operations on p2 are even glider number, and any odd number makes operations on p1 you will know that all your p2 intputs are of the same parity (thus keeping only 2 options).

One can modify "Schrodinger glue" to make unknown parity input glider only every second glider emission, and to limit the number of parallel option to two. This will probably yield the above trick anyway.
dvgrn wrote: Any plans for a two-week run this time around
I've updated the repository with depth 11. Depth 12 will probably take a week or so, and I don't think I'll go deeper (depth 11 is 170K and depth 12 is 600K).

EDIT I've modified the MonochromaticGlue script to improve performance, mainly needed for higher depth searches. Probably depth 13 now will take a week (it's 10 times faster now).

EDIT2 I've fixed a serious bug (the first recipe was not reported). If you took previous version, please update it now. I've also created WSSSearch script that reports all *WSS recipes (sroted by *WSS type) and edge shooters are reported separately. If you choose to use it make sure you update the destinationPath variable.

I'm actually considering to modify my script for monochromatic edge shooters and run it for few days.

Meanwhile I guess I'll use not optimized recipes to modify all 8 timing adjustments for your MWSS recipe.

If I would see that the monochromatic projects advances well, I might switch to general case skipping the D59, as there are little that can be used from that case for the general one. I'm considering to switch to D58 already, as the monochromatic recipes can be adopted to any D(2N).

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 26th, 2015, 10:26 am

I'm trying to build adjustable even step kit. For now here is forward 118 recipe (just replace the glider reflector to be odd using Jason's collection):

Code: Select all

x = 162, y = 252, rule = LifeHistory
19$98.C.C$101.C$97.C3.C$97.C3.C$101.C$98.C2.C$99.3C4$109.C.C$112.C$
108.C3.C$112.C$109.C2.C$110.3C54$38.2C$38.3C$38.3C$37.C.2C$37.3C$38.C
14$56.2C$56.3C$55.C.2C$43.2C10.3C$43.3C10.C$43.3C$43.3C$42.C.2C$42.3C
$43.C55$98.C.C$97.C$97.C3.C$97.C3.C$97.C18.2C$97.C2.C14.3C$97.3C15.3C
$115.3C$115.2C.C$116.3C$117.C39$92.C.C$95.C$91.C3.C$95.C$92.C2.C4.2C$
93.3C3.C2.C$99.C.C$100.C!
If I'm going to use recipes that were found by the latest WSSSearch script, LWSS will come out of the box, as I previously wanted to use only MWSS and HWSS (mainly for simplicity) but now I think it's completely not necessary.

EDIT Adjustable 118 backward:

Code: Select all

x = 247, y = 582, rule = LifeHistory
25$98.C.C$101.C$97.C3.C$97.C3.C$101.C$98.C2.C$99.3C212$172.C.C$171.C$
171.C3.C$171.C3.C$171.C$171.C2.C$171.3C23$170.C.C$173.C$169.C3.C$169.
C3.C$173.C$170.C2.C$171.3C5$156.2C$155.3C$155.3C$155.3C$155.2C.C$156.
3C$157.C76$141.2C$140.3C$140.2C.C$141.3C$142.C76$98.C.C$97.C$97.C3.C$
97.C3.C$97.C$97.C2.C$97.3C55$108.2C$108.3C$108.3C$107.C.2C$107.3C$
108.C7$87.C.C$86.C6.2C$86.C3.C.C2.C$86.C5.C.C$86.C2.C3.C$86.3C!
EDIT2 118 forward helix:

Code: Select all

x = 82, y = 145, rule = LifeHistory
6.C$5.C.C$5.C.C$6.C2$.2C7.2C$C2.C5.C2.C$.2C7.2C2$6.C$5.C.C$5.C.C$6.C
4$9.3C$8.C2.C$11.C$7.C3.C$11.C$8.C.C49$74.C$73.3C$72.2C.C$72.3C$72.3C
$72.3C$73.2C15$79.C$78.3C$78.C.2C$79.3C$79.3C$79.3C$79.2C42$5.3C$5.C
2.C$5.C$5.C3.C$5.C$6.C.C!
EDIT3 118 negative helix:

Code: Select all

x = 66, y = 366, rule = LifeHistory
63.2C$62.3C$62.3C$62.2C.C$63.3C$64.C177$.2C$3C$3C$3C$2C.C$.3C$2.C12$
10.C.C$9.C$9.C3.C$9.C$9.C2.C$9.3C150$58.C.C$57.C$57.C3.C$57.C$57.C2.C
$57.3C2$63.C$62.C.C$62.C.C$63.C!
Now only in-front deletion + adjustment script is left.

EDIT4 Collection of glider + *WSS -> SLs + glider (for the in front deletion):

Code: Select all

x = 538, y = 713, rule = LifeHistory
266.2C$126.2C137.3C148.2C$126.3C136.3C147.3C$126.3C136.2C.C146.3C$3.C
121.C.2C137.3C146.2C.C106.2C$4.C120.3C139.C148.3C105.3C$4.C121.C290.C
106.3C$4.C519.2C.C$.C2.C520.3C$2.3C521.C28$135.3C$135.C$136.C3$12.3C$
12.C522.3C$13.C521.C$536.C3$276.3C147.3C$276.C149.C$277.C149.C115$.C.
C$C125.2C$C3.C120.3C$C124.3C$C2.C121.2C.C$3C123.3C$11.C115.C$10.C$10.
3C9$137.C$136.C$136.3C130$137.C.C$140.C$140.C$9.2C126.C2.C$9.3C126.3C
$8.C.2C$8.3C263.C.C$9.C267.C$277.C$274.C2.C$275.3C15$19.3C$19.C$20.C
12$286.3C$286.C$287.C5$149.3C$149.C$150.C122$8.C.C$7.C$7.C3.C$7.C3.C$
7.C$7.C2.C$7.3C8$19.C$18.C$18.3C160$144.C.C$143.C$143.C3.C$143.C3.C$
143.C$143.C2.C$143.3C$11.C.C$10.C$10.C3.C300.C.C$10.C3.C299.C$10.C
303.C3.C$10.C2.C300.C3.C$10.3C301.C$314.C2.C$314.3C15$325.3C$325.C$
326.C14$21.3C$21.C$22.C2$154.3C$154.C$155.C!
Last edited by simsim314 on June 26th, 2015, 3:29 pm, edited 1 time in total.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 26th, 2015, 12:23 pm

simsim314 wrote:I'm actually considering to modify my script for monochromatic edge shooters and run it for few days.
Sounds good -- I'll be interested to see what shows up. Might be worth having a way to set up a different target population and/or bounding box for each search level.

For example, it's probably good to set limits to avoid pi-explosion-sized things early on... hey, those three P2 constellations in a pi explosion _could_ be Schroedingered away, no doubt. (But it seems awfully unlikely that it's worth the trouble!)

But especially at the very end of the recipe, it might not be so bad to have pi-explosion-sized leftovers, as long as the actual reaction was an edge-shooter *WSS. Cleaning up junk is pretty cheap, and lots of choices of new targets can make the next construction _really_ cheap.

-- Especially for the upship recipes. Isn't it true that the debris from those construction recipes can actually obstruct the upship lane, as long as the pre-trigger construction doesn't touch that lane? The debris can't extend as far as the _next_ upship lane to be constructed, but it would actually be good if it came as close as possible -- less distance to push junk.

The downship streams don't have that luxury, though. More debris might still be good, if it moves a reasonable distance toward the slow salvo, again to provide more target choices.

... Good Golly, what a strange design.

Meanwhile, it turns out that it's quite possible to adapt the new MWSS recipe to monochromatic P1. It takes one more glider for the Schroedinger cleanup, and the nice simple one-glider block pull turns into a 12-glider monstrosity (is there something cheaper? I built it by hand without asking for help from NewGlue.)

So we have a nice upper bound, at least: the recipe to beat is 21 gliders.

Code: Select all

#C 5,-3,13,-3,-3,5,-13,-1,9,11,19,-1,-1,-1,-1,-5,-7,-23,-19,-11,-17
#C (add NOP four places from the end if needed, between -7 and -23)
x = 1428, y = 1411, rule = LifeHistory
1426.2C$1426.2C63$1366.3A$1368.A$1367.A62$1294.3A$1296.A$1295.A62$
1246.3A$1248.A$1247.A62$1166.3A$1168.A$1167.A62$1102.3A$1104.A$1103.A
62$1046.3A$1048.A$1047.A62$964.3A$966.A$965.A62$912.3A$914.A$913.A62$
858.3A$860.A$859.A62$796.3A$798.A$797.A62$740.3A$742.A$741.A62$656.3A
$658.A$657.A62$592.3A$594.A$593.A62$528.3A$530.A$529.A62$464.3A$466.A
$465.A62$396.3A$398.A$397.A62$330.3A$332.A$331.A126$186.3A$188.A$187.
A62$126.3A$128.A$127.A62$70.3A$72.A$71.A62$3A$2.A$.A!
There's no way around collecting a bunch more edge-shooter recipes, though. Just retiming this recipe won't be enough any more, because in an even-step caterloopillar this recipe can't possibly build an odd-parity stream (i.e., an MWSS at a (0,1) offset from this recipe's output).

Could build the four seed SLs in a completely different way, I suppose, but then an additional color-changing mechanism would be needed for the trigger glider, in addition to the retiming toolkit. Could work, but it might be a bit expensive.

An interesting thing to try might be a version of the SL compiler script for an even-step helix, where one of the adjustable values is the timing between read heads.

Then put together a recipe that builds two successive downship streams... and see how the math works out for adjusting the relative locations of the two streams, without changing the recipe.

My current crackpot theory is that adding empty helices to the spaceship will change the *WSSes' relative locations very nicely for some read-head periods, but not for others -- seems as if there will be a few trouble spots where a lot more than eight different recipes/retiming tricks will be needed (?).

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: David Bell's engineless caterpillar idea revisited

Post by simsim314 » June 26th, 2015, 1:22 pm

A recipe of length 10 turned up almost exactly as yours (but with extra tub that doesn't change anything - only makes cleaning a bit harder):

Code: Select all

x = 1180, y = 1173, rule = B3/S23
2o$2o29$25b3o$25bo$26bo134$153b3o$153bo$154bo118$281b3o$281bo$282bo
124$409b3o$409bo$410bo118$537b3o$537bo$538bo130$665b3o$665bo$666bo124$
793b3o$793bo$794bo118$921b3o$921bo$922bo146$1049b3o$1049bo$1050bo110$
1177b3o$1177bo$1178bo!
The recipe is actually of length 11 (the first one simply adjusts the block lane):
6, 6, 14, 6, 4, -4, 0, -2, -10, 10, -6
dvgrn wrote:Might be worth having a way to set up a different target population and/or bounding box for each search level.
Maybe but from experience those magic parameters tend to work worse usually. Unless you have very precise formulas for them, based on some calculation, it's better to use less magic numbers.
dvgrn wrote:But especially at the very end of the recipe
My script posts any *WSS regardless of leftovers even p2 are accepted as long as some *WSS was emitted (in the right direction). For depth 11 out of ~170K options, only ~120 LWSS and 6 MWSS turned up, and no HWSS at all, and single edge shooter that I've posted (no LWSS edge shooter as well).

On the other hand, considering the fact that after two weeks of run I got nice collection of edge shooters, and the computational intensity is pretty similar in 3SL search and in Glue search, I can imagine that some edge shooters will come around after similar search time.
dvgrn wrote:There's no way around collecting a bunch more edge-shooter recipes, though.
Yep the even and the odd *WSS are separate creatures in even step design. Lets hope the search will yield some results and no GPU coding will be needed.
dvgrn wrote:My current crackpot theory is that adding empty helices to the spaceship will change the *WSSes' relative locations very nicely for some read-head periods, but not for others -- seems as if there will be a few trouble spots where a lot more than eight different recipes/retiming tricks will be needed (?).
I didn't get this point. My current belief is that 8 retimings and 2 *WSS parities is universal toolkit. BTW for similar reason I mentioned earlier, and to simplify our life by complicating to the worst case we must handle, one can choose to work only with periods dividable by 8. If we want to find speed of m/n we can use step 8m and period 8n. I think it makes it harder to work at first, on the other hand we must solve this case as well sometime, and any other case can be reduced to this one so we can chose to concentrate on it and only it.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: David Bell's engineless caterpillar idea revisited

Post by dvgrn » June 26th, 2015, 6:11 pm

simsim314 wrote:
dvgrn wrote:My current crackpot theory is that adding empty helices to the spaceship will change the *WSSes' relative locations very nicely for some read-head periods, but not for others -- seems as if there will be a few trouble spots where a lot more than eight different recipes/retiming tricks will be needed (?).
I didn't get this point. My current belief is that 8 retimings and 2 *WSS parities is universal toolkit.
Okay, I'll see if I can dig up one of the trouble spots, then. I may still be missing something completely obvious...!

The 59-step helices would be no problem, I think -- you can always get different *WSS output locations by leaving a few helix cycles empty.

Let's look at a random sample 118-step case. The downward read heads are following each other 512 cells apart, so the upward read heads will be (512-2*118) = 276 cells apart. Given the first *WSS stream, there are 276*4 = 1104 different possible relative placements for the second stream. Some are going to be more expensive than others. Here's a really cheap place for the second stream:

Code: Select all

x = 174, y = 2999, rule = LifeHistory
170.C.C$169.C$169.C3.C$169.C3.C$169.C$169.C2.C$169.3C23$168.C.C$171.C
$167.C3.C$167.C3.C$171.C$168.C2.C$169.3C5$154.2C$153.3C$153.3C$153.3C
$153.2C.C$154.3C$155.C76$139.2C$138.3C$138.2C.C$139.3C$140.C76$96.C.C
$95.C$95.C3.C$95.C3.C$95.C$95.C2.C$95.3C55$106.2C$106.3C$106.3C$105.C
.2C$105.3C$106.C7$85.C.C$84.C$84.C3.C$84.C$84.C2.C$84.3C14$96.C.C$99.
C$95.C3.C$95.C3.C$99.C$96.C2.C$97.3C212$170.C.C$169.C$169.C3.C$169.C
3.C$169.C$169.C2.C$169.3C23$168.C.C$171.C$167.C3.C$167.C3.C$171.C$
168.C2.C$169.3C5$154.2C$153.3C$153.3C$153.3C$153.2C.C$154.3C$155.C76$
139.2C$138.3C$138.2C.C$139.3C$140.C76$96.C.C$95.C$95.C3.C$95.C3.C$95.
C$95.C2.C$95.3C55$106.2C$106.3C$106.3C$105.C.2C$105.3C$106.C7$85.C.C$
84.C$84.C3.C$84.C$84.C2.C$84.3C14$96.C.C$99.C$95.C3.C$95.C3.C$99.C$
96.C2.C$97.3C212$170.C.C$169.C$169.C3.C$169.C3.C$169.C$169.C2.C$169.
3C23$168.C.C$171.C$167.C3.C$167.C3.C$171.C$168.C2.C$169.3C5$154.2C$
153.3C$153.3C$153.3C$153.2C.C$154.3C$155.C76$139.2C$138.3C$138.2C.C$
139.3C$140.C76$96.C.C$95.C$95.C3.C$95.C3.C$95.C$95.C2.C$95.3C55$106.
2C$106.3C$106.3C$105.C.2C$105.3C$106.C7$85.C.C$84.C$84.C3.C$84.C$84.C
2.C$84.3C14$96.C.C$99.C$95.C3.C$95.C3.C$99.C$96.C2.C$97.3C212$170.C.C
$169.C$169.C3.C$169.C3.C$169.C$169.C2.C$169.3C23$168.C.C$171.C$167.C
3.C$167.C3.C$171.C$168.C2.C$169.3C5$154.2C$153.3C$153.3C$153.3C$153.
2C.C$154.3C$155.C76$139.2C$138.3C$138.2C.C$139.3C$140.C76$96.C.C$95.C
$95.C3.C$95.C3.C$95.C$95.C2.C$95.3C55$106.2C$106.3C$106.3C$105.C.2C$
105.3C$106.C7$85.C.C$84.C$84.C3.C$84.C$84.C2.C$84.3C14$96.C.C$99.C$
95.C3.C$95.C3.C$99.C$96.C2.C$97.3C212$170.C.C$169.C$169.C3.C$169.C3.C
$169.C$169.C2.C$169.3C23$168.C.C$171.C$167.C3.C$167.C3.C$171.C$168.C
2.C$169.3C5$154.2C$153.3C$153.3C$153.3C$153.2C.C$154.3C$155.C76$139.
2C$138.3C$138.2C.C$139.3C$140.C76$96.C.C$95.C$95.C3.C$95.C3.C$95.C$
95.C2.C$95.3C55$106.2C$106.3C$106.3C$105.C.2C$105.3C$106.C$15.2A$14.A
2.A$14.A2.A$15.2A3$16.2A67.C.C$16.2A66.C$84.C3.C$7.2A75.C$6.A2.A74.C
2.C$7.2A75.3C3$11.2A3.2A$11.2A2.A2.A$15.A2.A$16.2A8$96.C.C$99.C$95.C
3.C$95.C3.C$99.C$96.C2.C$97.3C43$9.2A$8.A2.A$8.A2.A$9.2A3$10.2A$10.2A
2$.2A$A2.A$.2A3$5.2A3.2A$5.2A2.A2.A$9.A2.A$10.2A27$15.2A$14.A2.A$14.A
2.A$15.2A3$16.2A$16.2A2$7.2A$6.A2.A$7.2A3$11.2A3.2A$11.2A2.A2.A$15.A
2.A$16.2A57$9.2A$8.A2.A$8.A2.A$9.2A3$10.2A$10.2A2$.2A$A2.A$.2A3$5.2A
3.2A$5.2A2.A2.A$9.A2.A$10.2A27$15.2A$14.A2.A$14.A2.A$15.2A3$16.2A$16.
2A152.C.C$169.C$7.2A160.C3.C$6.A2.A159.C3.C$7.2A160.C$169.C2.C$169.3C
$11.2A3.2A$11.2A2.A2.A$15.A2.A$16.2A19$168.C.C$171.C$167.C3.C$167.C3.
C$171.C$168.C2.C$169.3C5$154.2C$153.3C$153.3C$153.3C$153.2C.C$154.3C$
155.C21$9.2A$8.A2.A$8.A2.A$9.2A3$10.2A$10.2A2$.2A$A2.A$.2A3$5.2A3.2A$
5.2A2.A2.A$9.A2.A$10.2A27$15.2A$14.A2.A$14.A2.A$15.2A3$16.2A$16.2A2$
7.2A$6.A2.A$7.2A130.2C$138.3C$138.2C.C$11.2A3.2A121.3C$11.2A2.A2.A
121.C$15.A2.A$16.2A57$9.2A$8.A2.A$8.A2.A$9.2A3$10.2A$10.2A2$.2A$A2.A$
.2A3$5.2A3.2A$5.2A2.A2.A$9.A2.A$10.2A84.C.C$95.C$95.C3.C$95.C3.C$95.C
$95.C2.C$95.3C21$15.2A$14.A2.A$14.A2.A$15.2A3$16.2A$16.2A2$7.2A$6.A2.
A$7.2A3$11.2A3.2A$11.2A2.A2.A$15.A2.A$16.2A17$106.2C$106.3C$106.3C$
105.C.2C$105.3C$106.C7$85.C.C$84.C6.2C$84.C3.C.C2.C$84.C5.C.C$84.C2.C
3.C$84.3C23$9.2A$8.A2.A$8.A2.A$9.2A3$10.2A$10.2A2$.2A$A2.A$.2A3$5.2A
3.2A$5.2A2.A2.A$9.A2.A$10.2A24$91.2C$90.C2.C$90.C.C$15.2A74.C$14.A2.A
$14.A2.A$15.2A3$16.2A$16.2A2$7.2A$6.A2.A$7.2A3$11.2A3.2A$11.2A2.A2.A$
15.A2.A$16.2A57$9.2A$8.A2.A$8.A2.A$9.2A3$10.2A$10.2A2$.2A$A2.A$.2A3$
5.2A3.2A$5.2A2.A2.A$9.A2.A$10.2A!
Now, suppose we want to put that second stream in a different place. If we have eight-or-so different MWSS recipes, that will give us seven more places to put MWSSes that are about equally cheap. That leaves 1096 other places where we might, if we're unlucky, have to place MWSSes.

As far I can see, the other way to adjust the relative MWSS placement location is by adding empty cycles to the helix. Here's the same pattern with one cycle and two cycles added -- i.e., the lowest loaf has moved down 118 and 236 steps:

Code: Select all

x = 646, y = 3453, rule = LifeHistory
108.C.C227.C.C227.C.C$111.C229.C229.C$107.C3.C225.C3.C225.C3.C$107.C
3.C225.C3.C225.C3.C$111.C229.C229.C$108.C2.C226.C2.C226.C2.C$109.3C
227.3C227.3C212$182.C.C227.C.C227.C.C$181.C229.C229.C$181.C3.C225.C3.
C225.C3.C$181.C3.C225.C3.C225.C3.C$181.C229.C229.C$181.C2.C226.C2.C
226.C2.C$181.3C227.3C227.3C23$180.C.C227.C.C227.C.C$183.C229.C229.C$
179.C3.C225.C3.C225.C3.C$179.C3.C225.C3.C225.C3.C$183.C229.C229.C$
180.C2.C226.C2.C226.C2.C$181.3C227.3C227.3C5$166.2C228.2C228.2C$165.
3C227.3C227.3C$165.3C227.3C227.3C$165.3C227.3C227.3C$165.2C.C226.2C.C
226.2C.C$166.3C227.3C227.3C$167.C229.C229.C76$151.2C228.2C228.2C$150.
3C227.3C227.3C$150.2C.C226.2C.C226.2C.C$151.3C227.3C227.3C$152.C229.C
229.C76$108.C.C227.C.C227.C.C$107.C229.C229.C$107.C3.C225.C3.C225.C3.
C$107.C3.C225.C3.C225.C3.C$107.C229.C229.C$107.C2.C226.C2.C226.C2.C$
107.3C227.3C227.3C55$118.2C228.2C228.2C$118.3C227.3C227.3C$118.3C227.
3C227.3C$117.C.2C226.C.2C226.C.2C$117.3C227.3C227.3C$118.C229.C229.C
7$97.C.C227.C.C227.C.C$96.C229.C229.C$96.C3.C225.C3.C225.C3.C$96.C
229.C229.C$96.C2.C226.C2.C226.C2.C$96.3C227.3C227.3C14$108.C.C227.C.C
227.C.C$111.C229.C229.C$107.C3.C225.C3.C225.C3.C$107.C3.C225.C3.C225.
C3.C$111.C229.C229.C$108.C2.C226.C2.C226.C2.C$109.3C227.3C227.3C212$
182.C.C227.C.C227.C.C$181.C229.C229.C$181.C3.C225.C3.C225.C3.C$181.C
3.C225.C3.C225.C3.C$181.C229.C229.C$181.C2.C226.C2.C226.C2.C$181.3C
227.3C227.3C23$180.C.C227.C.C227.C.C$183.C229.C229.C$179.C3.C225.C3.C
225.C3.C$179.C3.C225.C3.C225.C3.C$183.C229.C229.C$180.C2.C226.C2.C
226.C2.C$181.3C227.3C227.3C5$166.2C228.2C228.2C$165.3C227.3C227.3C$
165.3C227.3C227.3C$165.3C227.3C227.3C$165.2C.C226.2C.C226.2C.C$166.3C
227.3C227.3C$167.C229.C229.C76$151.2C228.2C228.2C$150.3C227.3C227.3C$
150.2C.C226.2C.C226.2C.C$151.3C227.3C227.3C$152.C229.C229.C76$108.C.C
227.C.C227.C.C$107.C229.C229.C$107.C3.C225.C3.C225.C3.C$107.C3.C225.C
3.C225.C3.C$107.C229.C229.C$107.C2.C226.C2.C226.C2.C$107.3C227.3C227.
3C55$118.2C228.2C228.2C$118.3C227.3C227.3C$118.3C227.3C227.3C$117.C.
2C226.C.2C226.C.2C$117.3C227.3C227.3C$118.C229.C229.C7$97.C.C227.C.C
227.C.C$96.C229.C229.C$96.C3.C225.C3.C225.C3.C$96.C229.C229.C$96.C2.C
226.C2.C226.C2.C$96.3C227.3C227.3C14$108.C.C227.C.C227.C.C$111.C229.C
229.C$107.C3.C225.C3.C225.C3.C$107.C3.C225.C3.C225.C3.C$111.C229.C
229.C$108.C2.C226.C2.C226.C2.C$109.3C227.3C227.3C212$182.C.C227.C.C
227.C.C$181.C229.C229.C$181.C3.C225.C3.C225.C3.C$181.C3.C225.C3.C225.
C3.C$181.C229.C229.C$181.C2.C226.C2.C226.C2.C$181.3C227.3C227.3C23$
180.C.C227.C.C227.C.C$183.C229.C229.C$179.C3.C225.C3.C225.C3.C$179.C
3.C225.C3.C225.C3.C$183.C229.C229.C$180.C2.C226.C2.C226.C2.C$181.3C
227.3C227.3C5$166.2C228.2C228.2C$165.3C227.3C227.3C$165.3C227.3C227.
3C$165.3C227.3C227.3C$165.2C.C226.2C.C226.2C.C$166.3C227.3C227.3C$
167.C229.C229.C76$151.2C228.2C228.2C$150.3C227.3C227.3C$150.2C.C226.
2C.C226.2C.C$151.3C227.3C227.3C$152.C229.C229.C76$108.C.C227.C.C227.C
.C$107.C229.C229.C$107.C3.C225.C3.C225.C3.C$107.C3.C225.C3.C225.C3.C$
107.C229.C229.C$107.C2.C226.C2.C226.C2.C$107.3C227.3C227.3C55$118.2C
228.2C228.2C$118.3C227.3C227.3C$118.3C227.3C227.3C$117.C.2C226.C.2C
226.C.2C$117.3C227.3C227.3C$118.C229.C229.C7$97.C.C227.C.C227.C.C$96.
C229.C229.C$96.C3.C225.C3.C225.C3.C$96.C229.C229.C$96.C2.C226.C2.C
226.C2.C$96.3C227.3C227.3C14$108.C.C227.C.C227.C.C$111.C229.C229.C$
107.C3.C225.C3.C225.C3.C$107.C3.C225.C3.C225.C3.C$111.C229.C229.C$
108.C2.C226.C2.C226.C2.C$109.3C227.3C227.3C212$182.C.C227.C.C227.C.C$
181.C229.C229.C$181.C3.C225.C3.C225.C3.C$181.C3.C225.C3.C225.C3.C$
181.C229.C229.C$181.C2.C226.C2.C226.C2.C$181.3C227.3C227.3C23$180.C.C
227.C.C227.C.C$183.C229.C229.C$179.C3.C225.C3.C225.C3.C$179.C3.C225.C
3.C225.C3.C$183.C229.C229.C$180.C2.C226.C2.C226.C2.C$181.3C227.3C227.
3C5$166.2C228.2C228.2C$165.3C227.3C227.3C$165.3C227.3C227.3C$165.3C
227.3C227.3C$165.2C.C226.2C.C226.2C.C$166.3C227.3C227.3C$167.C229.C
229.C76$151.2C228.2C228.2C$150.3C227.3C227.3C$150.2C.C226.2C.C226.2C.
C$151.3C227.3C227.3C$152.C229.C229.C76$108.C.C227.C.C227.C.C$107.C
229.C229.C$107.C3.C225.C3.C225.C3.C$107.C3.C225.C3.C225.C3.C$107.C
229.C229.C$107.C2.C226.C2.C226.C2.C$107.3C227.3C227.3C55$118.2C228.2C
228.2C$118.3C227.3C227.3C$118.3C227.3C227.3C$117.C.2C226.C.2C226.C.2C
$117.3C227.3C227.3C$118.C229.C229.C7$97.C.C227.C.C227.C.C$96.C229.C
229.C$96.C3.C225.C3.C225.C3.C$96.C229.C229.C$96.C2.C226.C2.C226.C2.C$
96.3C227.3C227.3C14$108.C.C227.C.C227.C.C$111.C229.C229.C$107.C3.C
225.C3.C225.C3.C$107.C3.C225.C3.C225.C3.C$111.C229.C229.C$108.C2.C
226.C2.C226.C2.C$109.3C227.3C227.3C212$182.C.C227.C.C227.C.C$181.C
229.C229.C$181.C3.C225.C3.C225.C3.C$181.C3.C225.C3.C225.C3.C$181.C
229.C229.C$181.C2.C226.C2.C226.C2.C$181.3C227.3C227.3C23$180.C.C227.C
.C227.C.C$183.C229.C229.C$179.C3.C225.C3.C225.C3.C$179.C3.C225.C3.C
225.C3.C$183.C229.C229.C$180.C2.C226.C2.C226.C2.C$181.3C227.3C227.3C
5$166.2C228.2C228.2C$165.3C227.3C227.3C$165.3C227.3C227.3C$165.3C227.
3C227.3C$165.2C.C226.2C.C226.2C.C$166.3C227.3C227.3C$167.C229.C229.C
76$151.2C228.2C228.2C$150.3C227.3C227.3C$150.2C.C226.2C.C226.2C.C$
151.3C227.3C227.3C$152.C229.C229.C76$108.C.C227.C.C227.C.C$107.C229.C
229.C$107.C3.C225.C3.C225.C3.C$107.C3.C225.C3.C225.C3.C$107.C229.C
229.C$107.C2.C226.C2.C226.C2.C$107.3C227.3C227.3C55$118.2C228.2C228.
2C$118.3C227.3C227.3C$118.3C227.3C227.3C$117.C.2C226.C.2C226.C.2C$
117.3C227.3C227.3C$118.C229.C229.C7$28.2A67.C.C158.2A67.C.C158.2A67.C
.C$28.2A66.C161.2A66.C161.2A66.C$96.C3.C225.C3.C225.C3.C$19.2A75.C
152.2A75.C152.2A75.C$18.A2.A74.C2.C148.A2.A74.C2.C148.A2.A74.C2.C$19.
2A75.3C150.2A75.3C150.2A75.3C3$23.2A3.2A223.2A3.2A223.2A3.2A$23.2A2.A
2.A222.2A2.A2.A222.2A2.A2.A$27.A2.A226.A2.A226.A2.A$28.2A228.2A228.2A
8$108.C.C227.C.C227.C.C$111.C229.C229.C$107.C3.C225.C3.C225.C3.C$107.
C3.C225.C3.C225.C3.C$111.C229.C229.C$108.C2.C226.C2.C226.C2.C$109.3C
227.3C227.3C49$22.2A228.2A228.2A$22.2A228.2A228.2A2$13.2A228.2A228.2A
$12.A2.A226.A2.A226.A2.A$13.2A228.2A228.2A3$17.2A3.2A223.2A3.2A223.2A
3.2A$17.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$21.A2.A226.A2.A226.A2.A$22.
2A228.2A228.2A33$28.2A228.2A228.2A$28.2A228.2A228.2A2$19.2A228.2A228.
2A$18.A2.A226.A2.A226.A2.A$19.2A228.2A228.2A3$23.2A3.2A223.2A3.2A223.
2A3.2A$23.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$27.A2.A226.A2.A226.A2.A$
28.2A228.2A228.2A63$22.2A228.2A228.2A$22.2A228.2A228.2A2$13.2A228.2A
228.2A$12.A2.A226.A2.A226.A2.A$13.2A228.2A228.2A3$17.2A3.2A223.2A3.2A
223.2A3.2A$17.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$21.A2.A226.A2.A226.A2.
A$22.2A228.2A228.2A33$28.2A228.2A228.2A$28.2A152.C.C73.2A152.C.C73.2A
152.C.C$181.C229.C229.C$19.2A160.C3.C63.2A160.C3.C63.2A160.C3.C$18.A
2.A159.C3.C62.A2.A159.C3.C62.A2.A159.C3.C$19.2A160.C67.2A160.C67.2A
160.C$181.C2.C226.C2.C226.C2.C$181.3C227.3C227.3C$23.2A3.2A223.2A3.2A
223.2A3.2A$23.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$27.A2.A226.A2.A226.A2.
A$28.2A228.2A228.2A19$180.C.C227.C.C227.C.C$183.C229.C229.C$179.C3.C
225.C3.C225.C3.C$179.C3.C225.C3.C225.C3.C$183.C229.C229.C$180.C2.C
226.C2.C226.C2.C$181.3C227.3C227.3C5$166.2C228.2C228.2C$165.3C227.3C
227.3C$165.3C227.3C227.3C$165.3C227.3C227.3C$165.2C.C226.2C.C226.2C.C
$166.3C227.3C227.3C$167.C229.C229.C27$22.2A228.2A228.2A$22.2A228.2A
228.2A2$13.2A228.2A228.2A$12.A2.A226.A2.A226.A2.A$13.2A228.2A228.2A3$
17.2A3.2A223.2A3.2A223.2A3.2A$17.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$21.
A2.A226.A2.A226.A2.A$22.2A228.2A228.2A27$27.2A458.2A$26.A2.A456.A2.A$
26.A2.A456.A2.A$27.2A458.2A3$6.3D19.2A206.3D19.2A206.3D19.2A$6.D2.D
18.2A206.D2.D18.2A206.D2.D18.2A$6.D229.D229.D$6.D3.D8.2A215.D3.D8.2A
215.D3.D8.2A$6.D11.A2.A214.D11.A2.A214.D11.A2.A$7.D.D9.2A130.2C84.D.D
9.2A130.2C84.D.D9.2A130.2C$150.3C227.3C227.3C$150.2C.C226.2C.C226.2C.
C$23.2A3.2A121.3C99.2A3.2A121.3C99.2A3.2A121.3C$23.2A2.A2.A121.C100.
2A2.A2.A121.C100.2A2.A2.A121.C$27.A2.A226.A2.A226.A2.A$28.2A228.2A
228.2A63$22.2A228.2A206.3D19.2A$22.2A228.2A206.D2.D18.2A$460.D$13.2A
228.2A215.D3.D8.2A$12.A2.A226.A2.A214.D11.A2.A$13.2A228.2A216.D.D9.2A
3$17.2A3.2A223.2A3.2A223.2A3.2A$17.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$
21.A2.A226.A2.A226.A2.A$22.2A84.C.C141.2A84.C.C141.2A84.C.C$107.C229.
C229.C$107.C3.C225.C3.C225.C3.C$107.C3.C225.C3.C225.C3.C$107.C229.C
229.C$107.C2.C226.C2.C226.C2.C$107.3C227.3C227.3C21$257.2A$256.A2.A$
230.3D23.A2.A$230.D2.D23.2A$230.D$230.D3.D$28.2A200.D27.2A228.2A$28.
2A201.D.D24.2A228.2A2$19.2A228.2A228.2A$18.A2.A226.A2.A226.A2.A$19.2A
228.2A228.2A3$23.2A3.2A223.2A3.2A223.2A3.2A$23.2A2.A2.A222.2A2.A2.A
222.2A2.A2.A$27.A2.A226.A2.A226.A2.A$28.2A228.2A228.2A17$118.2C228.2C
228.2C$118.3C227.3C227.3C$118.3C227.3C227.3C$117.C.2C226.C.2C226.C.2C
$117.3C227.3C227.3C$118.C229.C229.C3$3D$D2.D$D$D3.D$D96.C.C227.C.C
227.C.C$.D.D92.C6.2C221.C6.2C221.C6.2C$96.C3.C.C2.C220.C3.C.C2.C220.C
3.C.C2.C$96.C5.C.C221.C5.C.C221.C5.C.C$96.C2.C3.C222.C2.C3.C222.C2.C
3.C$96.3C227.3C227.3C29$22.2A228.2A228.2A$22.2A228.2A228.2A2$13.2A
228.2A228.2A$12.A2.A226.A2.A226.A2.A$13.2A228.2A228.2A3$17.2A3.2A223.
2A3.2A223.2A3.2A$17.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$21.A2.A226.A2.A
226.A2.A$22.2A228.2A228.2A24$103.2C$102.C2.C$102.C.C$103.C383.2A$486.
A2.A$486.A2.A$487.2A3$28.2A228.2A228.2A$28.2A228.2A228.2A2$19.2A228.
2A228.2A$18.A2.A226.A2.A226.A2.A$19.2A228.2A228.2A3$23.2A3.2A223.2A3.
2A223.2A3.2A$23.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$27.A2.A226.A2.A226.A
2.A$28.2A228.2A228.2A63$22.2A228.2A228.2A$22.2A228.2A228.2A2$13.2A
228.2A228.2A$12.A2.A226.A2.A226.A2.A$13.2A228.2A228.2A3$17.2A3.2A223.
2A3.2A223.2A3.2A$17.2A2.A2.A222.2A2.A2.A222.2A2.A2.A$21.A2.A226.A2.A
226.A2.A$22.2A228.2A228.2A24$333.2C$332.C2.C$332.C.C$333.C6$258.2A
228.2A$258.2A228.2A2$249.2A228.2A$248.A2.A226.A2.A$249.2A228.2A3$253.
2A3.2A223.2A3.2A$253.2A2.A2.A222.2A2.A2.A$257.A2.A226.A2.A$258.2A228.
2A63$252.2A228.2A$252.2A228.2A2$243.2A228.2A$242.A2.A226.A2.A$243.2A
228.2A3$247.2A3.2A223.2A3.2A$247.2A2.A2.A222.2A2.A2.A$251.A2.A226.A2.
A$252.2A228.2A24$563.2C$562.C2.C$562.C.C$563.C6$488.2A$488.2A2$479.2A
$478.A2.A$479.2A3$483.2A3.2A$483.2A2.A2.A$487.A2.A$488.2A63$482.2A$
482.2A2$473.2A$472.A2.A$473.2A3$477.2A3.2A$477.2A2.A2.A$481.A2.A$482.
2A!
As the red-marked upships show, every time you add one empty cycle, you move the second MWSS up 40 cells relative to the first one.

That gives us all the adjustability we need. The GCD of 236 and 40 is 4, so I believe that means that for this speed and step combination, only 4 of each of the *WSS recipes will be needed.

-- But where does that 40 come from, exactly? Is there some speed and step combination where that 40 will change to something else?

Probably. I haven't done any more experimenting yet, and I'm out of time for now... Further bulletins as events warrant.

The question is: is there a speed and step combination where that 40 changes to zero? If so, adding cycles stops being a source of variability. What other source of variability do we have, so that we won't need 1104 different MWSS recipes?

(Or at least ~20 different recipes, one for each upship stream we need to build. But those would be twenty very unlikely recipes.)

Tangential note: There's an interesting Doppler effect between the positive and negative helices, that I hadn't really thought about -- upship streams are closer together by 2*(step size) than downship streams. (Right?) Not sure exactly how that affects the math for the two halves yet.

Post Reply