Soup search results

[Alexey_Nigin]

3 trillion objects!

[Scorbie]

EDIT: Pardon, the last pattern is based on a different octagon. But you can still interbreed:

Code: Select all

x = 16, y = 18, rule = B3/S23
6b3o$7bo2$6b3o$5bo3bo$4bo5bo$3bo7bob2o$2obo7bo3bo$2obo7b2o$3bo7bobo$4b
o5bo2b2o$5bo3bo$6b3o2$5bo3bo$4b2o3b2o$4b2o3b2o$4bo5bo!

Nice to see that the octagon 4 and the octagon 3(?) can be interbred!

Here's the oscillator supported by a single octagon 4 quadrant:

Code: Select all

x = 12, y = 10, rule = B3/S23
7b2o$bo3b3o$o3bo4bobo$bo2b5ob2o$2o2b2o2bo$2o2b2o2bo$o3b5ob2o$4bo4bobo$
5b3o$7b2o!

EDIT: It can be attatched to an octagon "five":

Code: Select all

x = 20, y = 21, rule = B3/S23
3bo$3b3o$6bo3b2o$5b2o3bo$18b2o$7b5o6bo$6bo5bo3bobo$b2o2bo7bo2b2o$2bobo
9bo$3b2o3b2o4bo$o3bo3b2o4bo$b2obo9bo$4bo9bo$5bo7bo2b3o$6bo5bo4b3o$7b5o
2$6bo5bo$5b2o5b2o$5b2o5b2o$5bo7bo!

[gameoflifeboy]

Did anyone else see the one soup for ov_p29, which produces a variant of the p29 pre-pulsar shuttle?

Code: Select all

#C [[ THEME 4 ]]
x = 31, y = 31, rule = B3/S23
bobbobbobboobbobobboobbobbobbob$
oboobbbbbobbobbobbobbobbbbboobo$
boboobbbbobobbbobbbobobbbboobob$
boobobbobooobbbobbbooobobboboob$
obooobobooooooboboooooobobooobo$
bbbbbbooobobobbobbobobooobbbbbb$
bbbboooboobbbbooobbbboobooobbbb$
obbobobbbbbobobbbobobbbbbobobbo$
bbbbooobbobobobbbobobobbooobbbb$
boooobobobooobbbbboooboboboooob$
obbooobbbobbbobobobbbobbbooobbo$
obooobbooobboobbboobbooobbooobo$
bobboobbbobobooooobobobbboobbob$
bbbbobbooboooobobooooboobbobbbb$
obbbbbobbbbbobooobobbbbbobbbbbo$
boooooobbbobooooooobobbboooooob$
obbbbbobbbbbobooobobbbbbobbbbbo$
bbbbobbooboooobobooooboobbobbbb$
bobboobbbobobooooobobobbboobbob$
obooobbooobboobbboobbooobbooobo$
obbooobbbobbbobobobbbobbbooobbo$
boooobobobooobbbbboooboboboooob$
bbbbooobbobobobbbobobobbooobbbb$
obbobobbbbbobobbbobobbbbbobobbo$
bbbboooboobbbbooobbbboobooobbbb$
bbbbbbooobobobbobbobobooobbbbbb$
obooobobooooooboboooooobobooobo$
boobobbobooobbbobbbooobobboboob$
boboobbbbobobbbobbbobobbbboobob$
oboobbbbbobbobbobbobbobbbbboobo$
bobbobbobboobbobobboobbobbobbob!

[dvgrn]

This asymmetric soup for an unusual eater2 variant takes several dozen ticks to stabilize, includes recognizable objects at various stages -- a B-heptomino, a glider, a bookend -- and ends up with the eater2 at the edge of its reaction envelope:

Code: Select all

#C 48-tick predecessor of diagonally symmetric eater2 variant
x = 22, y = 14, rule = LifeHistory
10.3A3.A$2.3A4.A.2A3.2A$7.A.2A6.2A$7.A.A6.2A$2A4.A.A11.2A$A4.A.A12.2A
$A2.A.A2.A3.2A$8.A3.2A$.3A.A$6.A$.3A$2.A$2.A$2.A!

Unfortunately it seems like a fairly intractable reaction for slow-salvo synthesis, or even for any kind of glider synthesis. No recipe is listed in Mark Niemiec's new database.

Looking at the other options in Catagolue, I think the other C1 soup for this object has much more potential. Maybe someone has already boiled this down to a standard glider recipe, and I missed it? It's nice having the Catagolue comments as a place to put these kinds of notes:

Code: Select all

#C cleanup method not meant to be taken seriously.
x = 72, y = 83, rule = LifeHistory
6.A.A$7.2A$7.A$62.A$61.A.A$62.2A18$66.A$65.A$65.A$65.2A$70.A$69.A.A$
69.A.A$70.A9$67.2A$66.2A$67.A16$66.2A$65.A.A$66.A21$2A$.2A$A!

[praosylen]

53-point soup:

Code: Select all

x = 16, y = 16, rule = B3/S23
ooobbobobbbobooo$
bbbooboobboobobo$
bbboobbboooobboo$
obboooooboobboob$
ooooobbooobboboo$
bbobbobooboobboo$
obboobooobobobob$
bboboooobbbbobbb$
bobbbbbbbbbbobbo$
bobbbboboobboboo$
obbbooboboobbobb$
oobobbobbobooobo$
oobooobooobooobb$
bbbbboooobbboooo$
obboobooobobbooo$
bboobbobbbobobbo!

[mniemiec]

From http://catagolue.appspot.com/object/xp2 ... zw23/b3s23:
This object currently takes 27 gliders, but hopefully from the following predecessor gleaned from the above (forms of which occur in multiple soups), hopefully this can be substantially reduced. I haven't been able to get the top glider and the R-pentomino to arrive simultaneously; I'll leave that to others who are more adept at working with methuselah predecessors at close quarters. The first soup on the above list uses something larger and different from the glider for the same effect, but synthesizing that would likely be more expensive.

Code: Select all

x = 48, y = 51, rule = B3/S23
45boo$45bobo$46bo$40b3o$6b3o3bo21bo$5bobbobbobo18bobo$4bo3boboobo17bo
4bo$4bo7bo17bob5o$4bo3bo7bo14bo$5bobbo5boo16b3o$6b3o6boo17bo17$3o$bbo$
bo11$15boo3boo$15boboboo$16bo4bo$10b3o$4bo29bo$bbobo27bobo$bo4bo24bo4b
o$ob5o23bob5o$bo29bo$bb3o27b3o$4bo29bo!

[gameoflifeboy]

There's been a natural occurrence of a new asymmetric P15 oscillator, the "Pulsar on pentadecathlon 2". Except it occurred in a symmetric soup:

Code: Select all

x = 31, y = 32, rule = B3/S23
boobbbobbbbooobbbbbbbbbbbbbbbbb$
oobobboobobbboobbbbbbbbbbbbbbbb$
bobboobobbobbbbobbbbbbbbbbbbbbb$
bbbboobboooboobbbbbbbbbbbbbbbbb$
obobbooobbbbobbobbbbbbbbbbbbbbb$
bobboooooboooooobbbbbbbbbbbbbbb$
bobobooobboobobbbbbbbbbbbbbbbbb$
boooboobobbooobobbbbbbbbbbbbbbb$
ooobbbboobbbboobbbbbbbbbbbbbbbb$
boooooboboboobobbbbbbbbbbbbbbbb$
oboboobooooooboobbbbbbbbbbbbbbb$
oboobobbobbbbbobbbbbbbbbbbbbbbb$
ooooboooboooooobbbbbbbbbbbbbbbb$
boooobbbobbbobbbbbbbbbbbbbbbbbb$
oobbbobboobbbbbbbbbbbbbbbbbbbbb$
obbbooobbboobbobbbbbbbbbbbbbbbb$
bbbbbbbbbbbbbbbbobboobbbooobbbo$
bbbbbbbbbbbbbbbbbbbbboobbobbboo$
bbbbbbbbbbbbbbbbbbobbbobbboooob$
bbbbbbbbbbbbbbbbooooooboooboooo$
bbbbbbbbbbbbbbbbobbbbbobboboobo$
bbbbbbbbbbbbbbboobooooooboobobo$
bbbbbbbbbbbbbbbboboobobobooooob$
bbbbbbbbbbbbbbbboobbbboobbbbooo$
bbbbbbbbbbbbbbbobooobboboobooob$
bbbbbbbbbbbbbbbbboboobbooobobob$
bbbbbbbbbbbbbbboooooobooooobbob$
bbbbbbbbbbbbbbbobbobbbbooobbobo$
bbbbbbbbbbbbbbbbboobooobboobbbb$
bbbbbbbbbbbbbbbobbbbobboboobbob$
bbbbbbbbbbbbbbbboobbboboobboboo$
bbbbbbbbbbbbbbbbbooobbbbobbboob!

[dvgrn]

The newest eater2 variant reported by the Twitterbot comes from a soup that definitely implies a glider synthesis, though I'm not sure exactly where the best attack point is:

Code: Select all

x = 37, y = 53, rule = B3/S23
bo$obo$2o4$11b2o$10bo2bo$11b2o4$8b3o$2b2o5bo2b2o$2b2o6bo$12b2o2$6b2o$
5bo2bo$4bo2bo$5bobo8$5bo$4bobo$4bobo$5bo7$7b2o$7b2o11$35bo$34b2o$34bob
o!

[Billabob]

Sorry if this is unhelpful, or even just irrelevant, but I have calculated that there is around a 6/10 chance that we will have a 13-snake by the time we reach 10^13 objects. I can give numbers on all of the snakes up to 31, but the higher the cell-count the more likely I am wrong.

EDIT: I got this number by looking at the total object count/occurences of each x-bit snake, made a graph in Excel, and predicted from there. Of course the data I used is no longer as accurate as it could be, as Catagolue has had more objects submitted... Is there a way to link my Excel sheet straight to the online data?

[mniemiec]

The newest eater2 variant reported by the Twitterbot comes from a soup that definitely implies a glider synthesis, though I'm not sure exactly where the best attack point is:

The top boat is unnecessary, and the bottom block can be replaced by a single glider. I wasted a lot of time trying to get all the pieces to fit, starting with a 3-glider R-pentomino. Once I dispensed with that preconception (this version makes it out of 5), all the other pieces fell easily into place, and can be either pre-built at par, or dynamically added at par. Three additional gliders are required for cleanup - two dynamically, and one after the fact, for a total of 18 gliders.

I had also considered synthesizing this using more traditional synthetic methods, i.e. starting from the 19-bit pseudo-still-life formed from three boats and a block in a ring, from 10 gliders, but since turning those boats into the three inductors would take more than 8 gliders, this idea is moot.

Code: Select all

x = 208, y = 80, rule = B3/S23
75bo$76bo$74b3o$$147bo$146bo$146b3o10$97bo$98boo$97boo$107bo$105boo$
106boo$$195bo$88bo71bo29bobboo$86bobo70bobo27bobobboo$87boo70bobo27bob
o$96bo63bo29bo$94bobo69boo28boo$95boo69boo28boo$90bobo$91boo$91bo10b3o
$102bo$25boo18boo18boo28boo6bo$3bo21bobo17bobo17bobo27bobo$bobo3bo18b
oo18boo18boo23b3obboo52boo3bo24boo3bo14boo3bo$bboobboo76boo7bo56bobobo
bo23bobobobo13bobobobo$6bobo74bobo6bo59bobobbo24bobobbo14bobobbo$bo83b
o65boob3o24boob3o14boob3o$oo$obo148boob3o24boob3o14boob3o$151boobobbo
23boobobbo13boobobbo$50bo105boo28boo18boo$49bo$49b3o14bo29bo$65bobo27b
obo$46b3o16bobo27bobo$46bo19bo29bo$47bo21$128boo$127boo$129bo5$123boo$
122boo$124bo!

[Extrementhusiast]

The newest eater2 variant reported by the Twitterbot comes from a soup that definitely implies a glider synthesis, though I'm not sure exactly where the best attack point is:

The top boat is unnecessary, and the bottom block can be replaced by a single glider. I wasted a lot of time trying to get all the pieces to fit, starting with a 3-glider R-pentomino. Once I dispensed with that preconception (this version makes it out of 5), all the other pieces fell easily into place, and can be either pre-built at par, or dynamically added at par. Three additional gliders are required for cleanup - two dynamically, and one after the fact, for a total of 18 gliders.

I had also considered synthesizing this using more traditional synthetic methods, i.e. starting from the 19-bit pseudo-still-life formed from three boats and a block in a ring, from 10 gliders, but since turning those boats into the three inductors would take more than 8 gliders, this idea is moot.

Code: Select all

RLE

You must have missed this one, listed right in your file:

Code: Select all

x = 77, y = 81, rule = B3/S23
obo$b2o$bo2$74bobo$74b2o$75bo9$24bo$25bo$23b3o$34bo$33bo$33b3o3$13bo$
14b2o$13b2o$21bo$22b2o$21b2o4$20bobo8bo$21b2o7b2o$21bo8bobo4$10b2o13b
2o$11b2o8bo2b2o$10bo10b2o3bo$20bobo4$23bo$22bobo$22bobo$23bo23$55b3o$
55bo$56bo5$50b3o$50bo$51bo!

[mniemiec]

I had also considered synthesizing this using more traditional synthetic methods, i.e. starting from the 19-bit pseudo-still-life formed from three boats and a block in a ring, from 10 gliders, but since turning those boats into the three inductors would take more than 8 gliders, this idea is moot.

You must have missed this one, listed right in your file:

I meant that THAT idea (i.e. the idea of synthesizing it the usual artificial way, from the pseudo-still-life) was moot, as the soup method that I posted is cheaper.

[Extrementhusiast]

I had also considered synthesizing this using more traditional synthetic methods, i.e. starting from the 19-bit pseudo-still-life formed from three boats and a block in a ring, from 10 gliders, but since turning those boats into the three inductors would take more than 8 gliders, this idea is moot.

You must have missed this one, listed right in your file:

I meant that THAT idea (i.e. the idea of synthesizing it the usual artificial way, from the pseudo-still-life) was moot, as the soup method that I posted is cheaper.

I meant my comment in reference to this part of your post (which I should have probably trimmed the former quote down to):

I wasted a lot of time trying to get all the pieces to fit, starting with a 3-glider R-pentomino. Once I dispensed with that preconception (this version makes it out of 5), all the other pieces fell easily into place, and can be either pre-built at par, or dynamically added at par.

I was noting that you had missed a three-glider method for synthesizing that R-pentomino, which was in your collection.

[mniemiec]

I was noting that you had missed a three-glider method for synthesizing that R-pentomino, which was in your collection.

Oh DUH! I looked at the synthesis you posted, carefully examined everything on the periphery and saw no changes, and totally overlooked the R-pentomino synthesis (because I had misinterpreted the comment). Now fixed. Thanks!

[Alexey_Nigin]

Some time ago I left a comment on Catagolue regarding this, and I think this pattern is worth posting here:

Code: Select all

x = 32, y = 32, rule = B3/S23
boobbbobbbobbobbbbobbobbbobbboob$
oobooboobbbbooobbooobbbboobooboo$
obbobbooobbbboobboobbbbooobbobbo$
booobooobboobboooobboobbooobooob$
bobboobbboobbbboobbbboobbboobbob$
bbbooooboobbbobbbbobbbooboooobbb$
oooobooobobbbbboobbbbboboooboooo$
booobbobboobooboobooboobbobbooob$
bbobbobboboooobooboooobobbobbobb$
bbbboooobbboboooooobobbboooobbbb$
obboobboobboboobboobobboobboobbo$
bbbobbbbooobboobboobbooobbbbobbb$
bobbbbboobbbbboooobbbbboobbbbbob$
ooobboboooooboboobobooooobobbooo$
booobbbbboooobooooboooobbbbbooob$
bbbooboooobboooooooobbooooboobbb$
bbbooboooobboooooooobbooooboobbb$
booobbbbboooobooooboooobbbbbooob$
ooobboboooooboboobobooooobobbooo$
bobbbbboobbbbboooobbbbboobbbbbob$
bbbobbbbooobboobboobbooobbbbobbb$
obboobboobboboobboobobboobboobbo$
bbbboooobbboboooooobobbboooobbbb$
bbobbobboboooobooboooobobbobbobb$
booobbobboobooboobooboobbobbooob$
oooobooobobbbbboobbbbboboooboooo$
bbbooooboobbbobbbbobbbooboooobbb$
bobboobbboobbbboobbbboobbboobbob$
booobooobboobboooobboobbooobooob$
obbobbooobbbboobboobbbbooobbobbo$
oobooboobbbbooobbooobbbboobooboo$
boobbbobbbobbobbbbobbobbbobbboob!

Link to Catagolue.

Run this pattern until it stabilises. You will like it.

(Also, 10000th post on Patterns board.)

[gmc_nxtman]

There seem to be an enormous number of oversized pseudo-objects being counted as strict objects.

[Alexey_Nigin]

There seem to be an enormous number of oversized pseudo-objects being counted as strict objects.

If an object is oversized, apgsearch just puts an "OVERSIZED" label and does nothing else. In particular, it doesn't check whether the object is strict or not. So I don't see any unexpected behaviour.

Of course, it would be very convenient if the limit was a bit larger, say 100 by 100 instead of 40 by 40.

[praosylen]

The first (in my opinion) new interesting object in B3/S23 C1 in quite a while:

Code: Select all

x = 16, y = 16, rule = B3/S23
obbooobooboboooo$
booobbobboooooob$
ooooobbobbboboob$
bobobbobbbooboob$
obbboobboobooobb$
boboobooobbooooo$
obbooobobobbbbbo$
oobbbobbbbbobbbo$
bobobobbboboobbo$
oobobbobobbbbbbb$
bobooobboooobbbb$
oboobboobobbobbo$
bbbbbobooooboobb$
bbbobooooooooobo$
bbooobbobobbbbbo$
bbbbbbobbbbbbbbo!

[gmc_nxtman]

Large unusual period 6 oscillator:

Code: Select all

x = 16, y = 16, rule = B3/S23
3b2o6b2o$5bo4bo$4bo6bo$o3bo6bo3bo$ob2o8b2obo$bo12bo2$7b2o$7b2o2$bo12bo
$ob2o8b2obo$o3bo6bo3bo$4bo6bo$5bo4bo$3b2o6b2o!

Unusual predecessor of a p2:

Code: Select all

x = 8, y = 7, rule = B3/S23
2o4b2o$2bo2bo$bo4bo$bo4bo2$2o4b2o$bo4bo!

Another eater2 variant:

Code: Select all

x = 12, y = 13, rule = B3/S23
5bo$4b2o$4bo$7bo$b2obo$2o3b3o$5bobo$3bob2ob2o$7bo$7bo$10b2o$9b2o$11bo!

And, "Dr. Monstaa"s SL:

Code: Select all

x = 10, y = 5, rule = B3/S23
b3o2b3o$3bo2bo$2bo4bo$2bo4bo$2o6b2o!

[mniemiec]

Large almost-psuedo (pseudo-pseudo?) still life:

This quickly turns into a period-6 oscillator that I've never seen before.

Unusual predecessor of a p2:

This can be synthesized from 6 gliders, and this predecessor would likely take at least 9, so while pretty, it doesn't confer any advantage.

Another eater2 variant:

This is an interesting variant, but the predecessor seems like it would be very difficult to synthesize.

And, "Dr. Monstaa"s SL:

Nice. This one looks like it might actually be plausible!

[gmc_nxtman]

Large almost-pseudo (pseudo-pseudo?) still life:

This quickly turns into a period-6 oscillator that I've never seen before.

Whoops, my bad!

Here is an interesting (but probably once again useless) to a p3 that seems to be a hybrid of candlefrobra and candelabra oscillators:

Code: Select all

x = 32, y = 32, rule = B3/S23
2bobobo2b2o3bo2bo3b2o2bobobo$bo2b2ob2o3bob4obo3b2ob2o2bo$o2bo2b3obo10b
ob3o2bo2bo$2b2ob2o4b2o6b2o4b2ob2o$2o3bob4obo2b2o2bob4obo3b2o$bob4o6b2o
2b2o6b4obo$ob2obobo6b4o6bobob2obo$b2obobobob3obo2bob3obobobob2o$b2obo
2b2o2bob2o2b2obo2b2o2bob2o$o3bo5b2obob2obob2o5bo3bo$obobo2bob2o3bo2bo
3b2obo2bobobo$3bo3b3ob3ob2ob3ob3o3bo$bob2o2bo3b2obo2bob2o3bo2b2obo$5bo
2b2obobo4bobob2o2bo$2o3b4obobo6bobob4o3b2o$bo2bobo2bobo3b2o3bobo2bobo
2bo$bo2bobo2bobo3b2o3bobo2bobo2bo$2o3b4obobo6bobob4o3b2o$5bo2b2obobo4b
obob2o2bo$bob2o2bo3b2obo2bob2o3bo2b2obo$3bo3b3ob3ob2ob3ob3o3bo$obobo2b
ob2o3bo2bo3b2obo2bobobo$o3bo5b2obob2obob2o5bo3bo$b2obo2b2o2bob2o2b2obo
2b2o2bob2o$b2obobobob3obo2bob3obobobob2o$ob2obobo6b4o6bobob2obo$bob4o
6b2o2b2o6b4obo$2o3bob4obo2b2o2bob4obo3b2o$2b2ob2o4b2o6b2o4b2ob2o$o2bo
2b3obo10bob3o2bo2bo$bo2b2ob2o3bob4obo3b2ob2o2bo$2bobobo2b2o3bo2bo3b2o
2bobobo!

Another useless predecessor to ship on boat with long tail on ship, derived from a fleet predecessor + spark:

Code: Select all

x = 12, y = 14, rule = B3/S23
2$8b3o$bo6bo2bo$obobo3bo2bo$2bobo4b3o$b2o3$8b2o2$7b3o$7bo2bo$9bo!

[calcyman]

We're rapidly approaching 4 * 10^12 objects now. Amongst the recent results is a new p15 oscillator:

http://catagolue.appspot.com/census/b3s23/C1/xp15

[codeholic]

Block+BNG+B -> Trans-twin-bee shuttle:
https://twitter.com/conwaylife/status/6 ... 8713647105

Code: Select all

x = 16, y = 16, rule = B3/S23
o3bobobo5b2o$3ob3obob2o3bo$ob4o2b2ob4o$2bob2ob4ob4o$5bo5b4o$o3b2o2bo2b
2obo$7bobobobobo$bo2b5o2b2ob2o$2b3obobobo2bobo$bo2b2o2b3obo$2bo3b2o3bo
3bo$5o2b3ob2ob2o$ob2o3b3o2bo$bob5o3b2ob2o$3o8b3obo$bo2bobob3ob2o!

[mniemiec]

Here is an interesting (but probably once again useless) to a p3 that seems to be a hybrid of candlefrobra and candelabra oscillators:

This is just four copies of Bent Keys connected together in a ring.

Another useless predecessor to ship on boat with long tail on ship, derived from a fleet predecessor + spark:

That particular spark doesn't look easy to make, but its evolution a few generations later is quite common, and the same effect can be achieved by 2 gliders:

Code: Select all

x = 46, y = 19, rule = B3/S23
o$b2o$2o$40b2o$13bobo23bobo$13b2o23bobo$14bo2b3o18bo$17bo21bo4b2o$12b
3o3bo19b2o3bobo$12bo27b2o2bo$3b2o8bo26bob2o$4b2o$3bo$7bo$7b2o$6bobo$
17bo$16b2o$16bobo!

[BobShemyakin]

Here is an interesting (but probably once again useless) to a p3 that seems to be a hybrid of candlefrobra and candelabra oscillators:

This is just four copies of Bent Keys connected together in a ring.

Another useless predecessor to ship on boat with long tail on ship, derived from a fleet predecessor + spark:

That particular spark doesn't look easy to make, but its evolution a few generations later is quite common, and the same effect can be achieved by 2 gliders:

Code: Select all

x = 46, y = 19, rule = B3/S23
o$b2o$2o$40b2o$13bobo23bobo$13b2o23bobo$14bo2b3o18bo$17bo21bo4b2o$12b
3o3bo19b2o3bobo$12bo27b2o2bo$3b2o8bo26bob2o$4b2o$3bo$7bo$7b2o$6bobo$
17bo$16b2o$16bobo!

Excellent! You can continue:

Code: Select all

x = 50, y = 28, rule = B3/S23
9bobo$10b2o$10bo$19bobo$19b2o$20bo$24bo$22b2o$23b2o19b2obo$o42bo2b2o$b
2o39bobo3b2o$2o40b2o4bo$40b2o7bo$13bobo23bobo5bobo$13b2o23bobo5bobo$
14bo2b3o18bo7b2o$17bo8b2o11bo4b2o$12b3o3bo6b2o11b2o3bobo$12bo14bo12b2o
2bo$3b2o8bo26bob2o$4b2o$3bo$7bo$7b2o$6bobo$17bo$16b2o$16bobo!

Bob Shemyakin