ConwayLife.com

Matthias, What is the current status of the search? I somehow lost the overview. I understand that the width 19 search as well as the width 21 gutter search produced no ships, but promising partial results. Correct? The width-19 search was completed with nothing found. The width-21 gutter search ha...

#01 **
#02 ***
#03 *
#04 **
#05 **
#06 ***
#07 ***
#08 ***
#09 **
#10 **
#11 *
#12 **
#13 **
#14 *
#15 *
#16 **
#17 **
#18 **
#19 **
#20 *
#21 ***
#22 **
#23 **
#24 ***
#25 *
#26 **
#27 *

I still regularly get crashes on the original qfind-c8.cpp with multiple threads. I believe it may be because in function success() both srows and ssrows are grown with realloc() . However I think that memory is shared between the threads and it's possible that realloc() might move the memory block...

How can I eliminate duplicate results? If you are finding a large number of spaceships, you will probably want to use the 'x' command which prevents the program from printing spaceships during the deepening stage. This will eliminate many of the duplicate patterns, but not all of them. There is no ...

I recommend getting a compiler that supports openMP. Where can I find one? I don't know your particular setup. Maybe the compiler you use already supports openMP and you just haven't found the right settings. I personally use g++ from the GNU compiler collection. Installation instructions will vary...

Sorry for my noobiness at spaceship searches (I don't use search programs other than apgsearch), by why are spaceship searches done by width? Calcyman gave one good reason, but there is another. It seems that these long, thin searches are generally the most effective. That is, you can get long part...

LaundryPizza03 wrote: ↑

February 18th, 2020, 11:15 am

replaced all the instances of omp_get_thread_num() with 1.

I think that you would need to replace them with 0, not 1. However, I recommend getting a compiler that supports openMP.

A weird thing happened when searching for a 2c/6o. A c/3 (unsimplified) popped out. If you want to find a period-6 ship, qfind might be better than zfind. Edit: If you want to find a symmetric 2c/6 ship of width 17, I recommend running qfind with the following settings: ./qfind B2n3/S23-q p6 k2 w9 ...

Sadly I don't currently have time to figure out why this is happening.

calcyman wrote: ↑

February 10th, 2020, 3:51 pm

lack of stator

I didn't even think about that. This is the first strictly volatile p16.

A for Awesome noticed that rliston's G1 searching turned up something new today: this new p16 oscillator. It breaks both the minimum population and bounding box records for a p16 oscillator. It has a few weak sparks, so it will probably give some smaller LCM oscillators, although I'm not sure if it...

Here is a collection of dump files and a specialized version of qfind for extending them using a w10 gutter search. The dump files were constructed from the longest partial from each report. They represent the points at which the partial changes from odd-symmetric to gutter symmetric. c8-extend.zip ...

if the search doesn't find a spaceship I'll be able to say that there "probably" isn't a width 12 spaceship. I have two more seedcolumns to try first though. I'm somewhat surprised that you haven't found a spaceship yet. Am I right in thinking that the last seed columns generally represent a larger...

Using one of the last 15-bitters to solve the last 19-bitter? Sheer poetry.

I didn't see a perfect match for the first 25 or 26 lines of the three longest width-19 gutter partials, but I was only hunting for the first of these eight phases... Am I missing an exact match somewhere in those early rows, or is that partial just being reported in a different phase, or something...

Clarification: this is not just because the dump files contain a couple of extra lines, lines 2 and 3, which could presumably be edited out? Actually, that's exactly the reason. Edit those lines out and you have a dump file that is valid for qfind-split. Running again with 256 nodes per piece gave ...

Sorry Dave. The process is really more involved than I made it sound. I've made a script (just an edit of qfind) to split a given dump file into pieces with a specified number of queue nodes per piece. qfind-split.zip compile with g++ qfind-split.cpp -O3 -o qsplit To use the script, run ./qsplit FF ...

I don't think we should hold voting on the wiki, since there are fewer people with wiki accounts and we have to give them permission to edit. I don't think we should have any more barriers to voting than we already do. As for the ease of adding descriptions, I think I've already done most of the leg...

I've tried to write summaries of all of the nominees: EDIT by dvgrn: had to remove the numbers, because the items are in a different order from testitemqlstudio's list in the LifeWiki article, and therefore different from the numbers in the voting thread -- seemed like leaving them in here would be ...

Here's a proposed description of that entry:

• Systematic survey of small patterns (by Nick Gotts): a systematic survey of all patterns up to 9 cells, as well as a survey of all single-cluster 10-cell patterns resulting in the discovery of a new longest-lived 10-cell methuselah, bunnies 10b.

Slightly better arrangement of Gosper guns in the p390 design Nice work fitting those queen bee stabilizations together. Here is a stator minimization that does not affect the bounding box: x = 48, y = 40, rule = LifeHistory 16.B4.5BA$14.11B4A$14.10B2AB4A2.B.B2A3B$13.2A8B3AB2A3BA3BA2BA3B$13.2A B.7B...

Yes, I notice that all the nominees are not into the list. I suppose there must be some good reasons for that, but I'm not sure I fully understand which ones. I think it may have just been a mistake. For example, scholar has been independently nominated by several people but didn't make it onto the...

We need to summarize all of the nominees and make a final list to vote on. Note that some of the nominees have not made it into the list in the first post, so we need to make sure those are included as well. Here are my proposals for some entries: David Hilbert (by Luka Okanishi and Aidan F. Pierce)...

my biggest fear with regards to this search is that I didn't implement the "push" detection properly. If so, it could be throwing out valid search paths. I figured out why those erroneous "push" files were being created. I accidentally did the initial split incorrectly I've looked through everythin...

It's probably difficult for one person to summarize all the entries. Those who think they understand the finer points of a particular entry should try to give a short summary. For example, here is a summary of one of my nominations: New gun designs (by Tanner Jacobi): clever new gun components that ...