I know this thread hasn't seen much progress in a while; I haven't had time to really do anything with this project for the past several months and I likely won't be able to do much in the next several months.
First, the design for this caterpillar's frontend will be the one I posted on January 6th 2021 (viewtopic.php?f=2&t=2286&start=50#p118407). The reason I think we should stick with this design instead of the first design (see second diagram) is because the first design was already getting way too large and it was not completable.
Second, the RLE in the post I linked above contains a "kind-of" solution to a problem I had in the first frontend design that I almost completed (see attached file "WIP_Spaceship_4_1_(this_file_was_created_only_for_a_forums_post).mc"). The problem that was that for every LWSS lane that was synthesized, a larger triangle had to be created, thus expanding the size of the ship exponentially. The "kind-of" solution to this problem was to make the first triangle smaller so that every LWSS lane I synthesized would expand at a lesser rate, thus causing the ship to expand at a lesser rate overall as it was being build. The RLE shows how the upward LWSS lanes should be synthesized differently than the original design to make the triangle shape smaller.
Third, although there is a rough outline for the frontend in my January 6th diagram, there still needs to be a design for the very front of the frontend and the LWSS lanes (highlighted red) in the diagram. I addressed this problem in this post post as the "first problem". In short, the red lane of LWSSs in the January 6th diagram needs to be permeable to Northwest travelling gliders so the XWSSs in the brown lane can reflect the Northwest gliders to the Northeast direction. The now Northeast gliders will collide with the LWSS lanes to synthesize blocks for the block tracks that the caterpillar uses to move (dvgrn made a frontend that resembles what I'm talking about here in the second RLE of the post, but that design isn't permeable to Northwest gliders, so it can't be used for this caterpillar).
Fourth, during January I was stuck on building the Northwest glider rake to produce the following output (addressed in this post post as the "second problem".):
Code: Select all
x = 139, y = 33, rule = B3/S23 bo$2o$obo3$4b2o27bo$4bobo25b2o$4bo27bobo3$9bo26b2o27bo$8b2o26bobo25b2o $8bobo25bo27bobo3$41bo26b2o27bo$40b2o26bobo25b2o$40bobo25bo27bobo3$73b o26b2o27bo$72b2o26bobo25b2o$72bobo25bo27bobo3$105bo26b2o$104b2o26bobo$ 104bobo25bo3$137bo$136b2o$136bobo!
Fifth, I want to talk about the "third problem" addressed in this post. I have already tried solving this problem in these two posts. Some of the XWSS combinations I found are potential solutions to the third problem, but they require new XWSS reflector technology to be built for them to work. The "Fourth" paragraph in this post shows that I was already having trouble building LWSS reflector technology; I didn't want to have to build reflectors for MWSSs and HWSSs as well. I also didn't want to build MWSS and LWSS reflectors because I could not find any way to independently reflect MWSSs and HWSSs using gliders (more specifically, I couldn't find a way to independently reflect West travelling MWSSs or LWSSs into North travelling MWSSs or LWSSs by using Northwest travelling gliders). I tried to restrict solutions to this problem to 2 LWSSs per reflection for 2 reasons:
- Progress on a LWSS reflector had already begun, and I didn't want to spend more time building MWSS or LWSS reflectors
- 2 LWSSs was what I believed to be the minimum amount of LWSSs to reflect Northwest to Northeast gliders
To solve this problem under the above criteria, and to confirm/deny that 2 LWSSs was the minimum amount of LWSSs, I tried to use gencols to enumerate all possible combinations of 2 upward travelling LWSSs colliding with a Northwest travelling glider, then only making a list of the combinations that meet the following criteria:
- Leaves no ash behind after colliding
- Before collision: 2 upward LWSSs and 1 Northwest glider
- After collision: 1 Northeast glider and nothing else.
After the list is complete I would make a smaller list of collision combinations that work properly for the (34,7)c/156 velocity based on the list of all combinations that meet the above 3 criteria. My attempt to use gencols to "confirm/deny that 2 LWSSs was the minimum amount of LWSSs" (and to possibly solve this problem at all) was unsuccessful in the end.
Sixth, the reason I want to use AbhpzTa's rakes like I have been using before is because their rakes have the least amount of block lanes known, and because their rake technology is sufficient to synthesize any object needed for the caterpillar to sustain itself. If anyone can come up with a forerake, backrake, and rephaser rake like AbhpzTa's rakes that uses less than 6 lanes, please post it in this thread.
Seventh, I know a backend to this caterpillar wasn't mentioned very often. This is because at first I wanted to focus on completing the frontend before I can begin thinking about, designing, and building the backend. Later on I decided to work on a rough design of a backend for this caterpillar. My latest progress can be found here in the last RLE of the post: viewtopic.php?f=2&t=2286&start=25#p107850
Progress on the January 6th diagram that has been made so far:
- This post containing the January 6th diagram and the RLE for the LWSS reflector
- These two posts containing some potential solutions to the "third problem" addressed in this post.
- My progress on the backend from this post (the last RLE of the post)
Summary of open problems to this caterpillar that need to be solved in order to finish it (as of this post):
- First, second, and third problems addressed in this post (second to last paragraph).
-- These problems need to be solved in order to assemble the frontend
- Backend needs to be designed and built using the same block tracks as AbhpzTa's rakes here: viewtopic.php?f=2&t=2286#p67276
-- I have already made a little progress on the backend here: viewtopic.php?f=2&t=2286&start=25#p107850
-- For a while I had a gut feeling that the backend design could be much simpler than this one, but I haven't and still don't have a way to prove it.