Sphenocorona wrote:The bounding box can be easily reduced to some extent by replacing some of the edge snarks with a chain of multiple snarks. The use of the new R421 was probably total overkill in trying to delay the right side gliders, but it worked.
As a demonstration of new technology, it seems fine the way it is -- no need to worry about bounding boxes.
If a reduction in size or repeat time is wanted, it will probably make sense to really restart from scratch. Looking at that three-stage rebuild of a loaf, I don't know why I ever thought that that was a good idea... the initial highway-robber reaction is nice and quick, but recovery is bound to be painfully slow. I seem to recall it was supposed to be a proof-of-concept, just to show that it was possible to snag a glider from one lane and leave gliders in adjacent lanes untouched.
With all the Bellman/ptbsearch/catgl activity lately, it seems quite possible that someone could dig up a much better H->bait converter -- either a direct H-to-loaf, or any other constellation that can be generated cleanly at the very edge of its reaction envelope by a H-to-junk converter, and that collapses fairly cleanly into a 90-degree glider. Anyone have any likely candidates lying around?
There are so many H-to-junk converters that show up in conduit searches, that one of them is bound to be just what we want... the problem is figuring out how to be reasonably efficient at testing and discarding all the other thousands of candidates. It would be nice to have an organized H->junk converter collection set up; pretty much any converter where the junk is right at the edge of its reaction envelope, with no catalysts in the way, has a small but nonzero probability of being useful to somebody sometime.
Sphenocorona wrote:Also, maybe a new elementary conduit stamp collection is in order, featuring all the new conduits along with the old?
Definitely! An update of Calcyman's collection would make a lot of sense now, with more readable labels for the sections.
A rebuild of the Herschel-conduit pattern in Golly's Signal-Circuitry folder is also long overdue -- none of Guam's discoveries are in there yet, even. Maybe the two should be combined into one large stamp collection, with some kind of labeling or index for the Herschel conduits, marking the locations of all the intermediate objects. It's really useful to see examples of how each elementary conduit can actually be used.
If someone wants to undertake the project, I'll check in the completed stamp collection to Golly immediately...!