indestructible pattern

For general discussion about Conway's Game of Life.
User avatar
calcyman
Moderator
Posts: 2964
Joined: June 1st, 2009, 4:32 pm

Re: indestructible pattern

Post by calcyman » July 3rd, 2010, 7:18 am

Would if the "wall" being repaired were a 99.99999% slow-glider eater?
No, it's too fragile. The remaining 0.00001% would cause the entire thing to blow, and reaching 99.99999% results in an enormous configuration, tens of billions of cells on a side. The wall should be something closer to a sparse array of blocks, with some highway robbers to defend the weak points at the corners.
although, of course, the UCC used in the gemini may not be the best suited to this particular task
Agreed. Gemini is designed specifically for perpetual self-relocation, and nothing else. That's why it's so efficient -- my latest general-purpose UCC (as yet unpublished) would take about 10^16 generations to replicate. I only need to sort out the tape for self-replication, and that's mainly an automatic process. It uses some technology from Gemini, specifically the idea of colliding two perpendicular synchronised gliders.
What do you do with ill crystallographers? Take them to the mono-clinic!

p46beth
Posts: 345
Joined: June 18th, 2010, 8:18 pm

Re: indestructible pattern

Post by p46beth » July 3rd, 2010, 4:27 pm

calcyman wrote:
Would if the "wall" being repaired were a 99.99999% slow-glider eater?
No, it's too fragile. The remaining 0.00001% would cause the entire thing to blow, and reaching 99.99999% results in an enormous configuration, tens of billions of cells on a side. The wall should be something closer to a sparse array of blocks, with some highway robbers to defend the weak points at the corners.
Ah, I see. A wall of sparse blocks would need to be very thick in order to block every diagonal. The possible issue with that layout is that if a block in the outer-most row of blocks is destroyed, the universal constructor would need to send gliders through the entire layer, which would need to be impossible without a generous supply of reflectors on scattered throughout the wall, to get gliders fired from the inside around other blocks. But those reflectors would also be targets for gliders coming from outside the wall. Although, I suppose it would be possible for the UCC to construct all necassary reflectors after detecting that a block has been destroyed, then delete them after repairing the wall.
The other potential issue is in order to detect the absence of a block, it would need to send out signals to each block's location at regular intervals (making it a very high period oscillator). The possibility remains, then, of a glider from outside the wall coming just as a block on the outermost layer is being checked. The glider could disrupt the signal, and there would be a lot of possible debris left, making it difficult to program the UCC to be ready for every situation.

User avatar
calcyman
Moderator
Posts: 2964
Joined: June 1st, 2009, 4:32 pm

Re: indestructible pattern

Post by calcyman » July 4th, 2010, 6:41 am

which would need to be impossible without a generous supply of reflectors on scattered throughout the wall, to get gliders fired from the inside around other blocks.
The Universal Constructor could use a multi-jointed construction arm to reach around the blocks, instead of using reflectors.
The other potential issue is in order to detect the absence of a block, it would need to send out signals to each block's location at regular intervals (making it a very high period oscillator).
The idea is for the collision to initiate a chain reaction, sending gliders towards the UCC (in certain, predictable positions), which then trigger the repair mechanism. If we can guarantee that every incoming glider does eventually reach the UCC, the UCC can do the rest of the work, sensing to ascertain the lane of the incoming glider.

Of course, this configuration would itself be unwieldly large, and I am more interested in an existence proof than an explicit example.
What do you do with ill crystallographers? Take them to the mono-clinic!

p46beth
Posts: 345
Joined: June 18th, 2010, 8:18 pm

Re: indestructible pattern

Post by p46beth » July 4th, 2010, 11:51 am

The idea is for the collision to initiate a chain reaction, sending gliders towards the UCC (in certain, predictable positions), which then trigger the repair mechanism. If we can guarantee that every incoming glider does eventually reach the UCC, the UCC can do the rest of the work, sensing to ascertain the lane of the incoming glider.
Ah, in that case the wall would need to be a bit more than
The wall should be something closer to a sparse array of blocks, with some highway robbers to defend the weak points at the corners.
Since a glider can easily destroy a block while releasing very little potential signals.
Of course, this configuration would itself be unwieldly large, and I am more interested in an existence proof than an explicit example.
Agreed, an actual slow glider-proof eater would likely be to large to use.

p46beth
Posts: 345
Joined: June 18th, 2010, 8:18 pm

Re: indestructible pattern

Post by p46beth » January 31st, 2011, 10:09 pm

The starwars rule 345/2/4 has a wall which is impenetrable from the inside:

Code: Select all

x = 28, y = 28, rule = 345/2/4
2.A22.A$.3A2.A2.A2.A2.A2.A2.A2.3A$2A.22A.2A$.2A22.2A$2.A22.A$2.A22.A$
.2A22.2A$2.A22.A$2.A22.A$.2A22.2A$2.A22.A$2.A22.A$.2A22.2A$2.A22.A$2.
A22.A$.2A22.2A$2.A10.2A10.A$2.A22.A$.2A22.2A$2.A22.A$2.A22.A$.2A22.2A
$2.A22.A$2.A22.A$.2A22.2A$2A.22A.2A$.3A2.A2.A2.A2.A2.A2.A2.3A$2.A22.A
!
As found in the thread for 345/2/4.

This is similar to the box mentioned earlier in this thread for b2/s2. It seems feasible that life could have a similar structure, although certainly not an orthogonal line of white cells...though allowing diagonal sides could work if there was a way to stabilize the corners.

Batmanifestdestiny
Posts: 54
Joined: June 9th, 2010, 3:53 pm

Re: indestructible pattern

Post by Batmanifestdestiny » February 6th, 2011, 7:51 pm

137ben wrote:The starwars rule 345/2/4 has a wall which is impenetrable from the inside:

Code: Select all

x = 28, y = 28, rule = 345/2/4
2.A22.A$.3A2.A2.A2.A2.A2.A2.A2.3A$2A.22A.2A$.2A22.2A$2.A22.A$2.A22.A$
.2A22.2A$2.A22.A$2.A22.A$.2A22.2A$2.A22.A$2.A22.A$.2A22.2A$2.A22.A$2.
A22.A$.2A22.2A$2.A10.2A10.A$2.A22.A$.2A22.2A$2.A22.A$2.A22.A$.2A22.2A
$2.A22.A$2.A22.A$.2A22.2A$2A.22A.2A$.3A2.A2.A2.A2.A2.A2.A2.3A$2.A22.A
!
As found in the thread for 345/2/4.

This is similar to the box mentioned earlier in this thread for b2/s2. It seems feasible that life could have a similar structure, although certainly not an orthogonal line of white cells...though allowing diagonal sides could work if there was a way to stabilize the corners.
I just tried a diagonal line in a toroidal universe, and a glider running into it makes a rather large mess.

User avatar
Tropylium
Posts: 421
Joined: May 31st, 2011, 7:12 pm
Location: Finland

Re: indestructible pattern

Post by Tropylium » July 1st, 2011, 12:01 am

The T-tetromino replicator (found in a cylindrical Life universe of circumference 6) seems to make a fairly indestructible wall. Granted, it doesn't exactly stay put…

Edit: Nevermind, it doesn't defend very well against irregular superstring salvos (which in turn are easily block'd by a few, well, blocks).

Post Reply