Soup search results

[wwei23]

Well, a spaceship or infinite growth can't be an RRO, so what's the point of not separating those?

[MathAndCode]

How do you think are the blocks' histories supposed to touch?

All four of them are formed from the stairstep hexomino.

I probably should have been more specific, as in histories for periodic objects only.

So then shouldn't traffic lights be counted as one object?

The BLSE's periodic history envelopes touch, which mean they are the same object

Oh; you meant that the reaction envelopes overlap. I thought that you meant that the BLSEs were formed from the same soup. The former makes much more sense (although I still disagree with it).

which mean they use the same population increase period, which since they happen to be in perfect antiphase, is 72.

But why does Catagolue care about the population increase period and not the actual period or the modulus?

[dvgrn]

How do you think are the blocks' histories supposed to touch?

All four of them are formed from the stairstep hexomino.

I probably should have been more specific, as in histories for periodic objects only.

So then shouldn't traffic lights be counted as one object?

The BLSE's periodic history envelopes touch, which mean they are the same object

Oh; you meant that the reaction envelopes overlap. I thought that you meant that the BLSEs were formed from the same soup. The former makes much more sense (although I still disagree with it).

which mean they use the same population increase period, which since they happen to be in perfect antiphase, is 72.

But why does Catagolue care about the population increase period and not the actual period or the modulus?

All of this is pretty much just a matter of what is's practical for Catagolue to figure out and keep track of, by processing a soup's ash after it has run for a long time.

The component pieces of LOM blocks or traffic lights don't have anything to tie them together, once they've formed. If apgsearch had to go back and run through every step of a soup's evolution, and track exactly where every object comes from, then the census operations would be hopelessly slow instead of blindingly fast.

In other words, maybe in some soup there's a traffic light or set of LOM blocks that wasn't created from a TL or LOM predecessor, but it doesn't make any sense for apgsearch to obsess over that possibility. Similarly, objects are considered by Catagolue to be the same object basically if it's highly inconvenient for apgsearch to figure out how to separate them.

[MathAndCode]

All of this is pretty much just a matter of what is's practical for Catagolue to figure out and keep track of, by processing a soup's ash after it has run for a long time.

Why can't Catagolue process the ash in real time?

The component pieces of LOM blocks or traffic lights don't have anything to tie them together, once they've formed. If apgsearch had to go back and run through every step of a soup's evolution, and track exactly where every object comes from, then the census operations would be hopelessly slow instead of blindingly fast.

I was confused about what overlapping histories meant. I understand now.

[Ch91]

Second Schick engine by Rob Liston:

Code: Select all

x = 16, y = 16, rule = B3/S23
3o3b2o4bo2bo$4obo3bo2b2obo$bo4bo7b2o$b4obo4bobobo$4o2bo2bo2bobo$b3ob2o
bo3bobo$bob2o2b9o$2b3obobobo4bo$2o2bo2b2obo$obob2obo2b2ob2o$2o2b4o2bo
bo2bo$ob2ob3o2b2ob2o$o4bo2bobo2bobo$2bob3o2bob2obo$o2bo4b4obobo$4b6ob
2o2bo!

Rob seems to have a knack for finding interesting patterns in his soups.

[MathAndCode]

Rob seems to have a knack for finding interesting patterns in his soups.

That's because he has done the vast majority of the apgcode G1 searches in ConwayLife. (G1 is a pseudo-symmetry that is the same as C1 (no special symmetry) except that it doesn't waste time cataloguing common objects, so it has more time to find and catalogue unusual objects.)

[bubblegum]

Rob seems to have a knack for finding interesting patterns in his soups.

That's because he has done the vast majority of the apgcode G1 searches in ConwayLife. (G1 is a pseudo-symmetry that is the same as C1 (no special symmetry) except that it doesn't waste time cataloguing common objects, so it has more time to find and catalogue unusual objects.)

small bit, but G1 is a symmetry

[MathAndCode]

small bit, but G1 is a symmetry

I was thinking in a mathematical sense instead of a programming sense.

[bubblegum]

small bit, but G1 is a symmetry

I was thinking in a mathematical sense instead of a programming sense.

I meant when you referred to G1 as an apgcode

[MathAndCode]

I meant when you referred to G1 as an apgcode

Yes; I should have thought about it in that context instead of in a mathematical sense.

[calcyman]

Here's a backshooting Corderpuffer which takes 896381 generations to tunnel through the ash:

Code: Select all

x = 16, y = 16, rule = B3/S23
b4ob4o2bobo$2b2o4b3obo2bo$b2o2bobo2bo3b2o$b2o3b6o2b2o$o2bo3bo2b2o2b2o$
4b4o2bo2b2o$bob2o3b2obo2b2o$b2ob2ob2o2b5o$4bo2bobo2bo2bo$b4ob4ob2ob2o$
b2o2bobobob4o$2o2bo2b2ob4o$o2bobo2bo5b2o$obo2b5o3bo$obob2ob2obo2b3o$2b
obobo4b4o!

[wwei23]

Impressive. Found by rliston, I'm assuming?

[bubblegum]

Here's a backshooting Corderpuffer which takes 896381 generations to tunnel through the ash:

Code: Select all

x = 16, y = 16, rule = B3/S23
b4ob4o2bobo$2b2o4b3obo2bo$b2o2bobo2bo3b2o$b2o3b6o2b2o$o2bo3bo2b2o2b2o$
4b4o2bo2b2o$bob2o3b2obo2b2o$b2ob2ob2o2b5o$4bo2bobo2bo2bo$b4ob4ob2ob2o$
b2o2bobobob4o$2o2bo2b2ob4o$o2bobo2bo5b2o$obo2b5o3bo$obob2ob2obo2b3o$2b
obobo4b4o!

*note to those looking for a Cordership-style puffer: this is a combination of two GPSEs.

Impressive. Found by rliston, I'm assuming?

Probably, although the percentage of yl384s in C1 and G1 that have 1 occurrence from a soup beginning with boooo is apparently 0, so I can't verify.

[wwei23]

Probably, although the percentage of yl384s in C1 and G1 that have 1 occurrence from a soup beginning with boooo is apparently 0, so I can't verify.

Try zz_LINEAR or PATHOLOGICAL. The debris at the back would prevent this being marked as a yl384.

[MikeP]

Code: Select all

#CXRLE Pos=-6,-49 Gen=2858
x = 18, y = 99, rule = B3/S23:T62,99
5bo4bo3b2obo$6bob8obo$9b8o$8b2o4b2o$7b3o$7b2o$7b2ob2o$10bo$5b4obo$5bo
3bo$4bo$4bo5bo$5bo2b2o$6b2o15$9bo$8b3o$7b4o$6bo5b4o$3o2bo7b4o$5bo4bo3b
2obo$6bob8obo$9b8o$8b2o4b2o$7b3o$7b2o$7b2ob2o$10bo$5b4obo$5bo3bo$4bo$
4bo5bo$5bo2b2o$6b2o15$9bo$8b3o$7b4o$6bo5b4o$3o2bo7b4o$5bo4bo3b2obo$6bo
b8obo$9b8o$8b2o4b2o$7b3o$7b2o$7b2ob2o$10bo$5b4obo$5bo3bo$4bo$4bo5bo$5b
o2b2o$6b2o15$9bo$8b3o$7b4o$6bo5b4o$3o2bo7b4o!

From this soup:

Code: Select all

#CXRLE Pos=-15,-16
x = 31, y = 33, rule = B3/S23:T31,33
2obob3obo2bo2b2ob2o2b2ob2o2b2o$bo2bo3bo3bo2bob2ob2o2b2o4bo$2bob2obob2o
2b2o2b4o3bob3o$2b2o4bo2bo2b2o4bo2bo3bobo$b3obob2o2b2o2b2obobo2bobo4bo$
bobo2b5o2b2obo2b4o7bo$obo2b2ob3ob4o3b3o6b2o$6o6bo2bo2b2o3bo2bob2o$2bob
4o2bob2o2b2o2b2ob3obo$4o2b3ob2o2b3o2b5o4bo$b3ob9o3bo2bo4b2ob2o$bo2bo2b
obob7obob3o2bobobo$bo5b2obob3o2b2o3bo4bobo$6o7b2ob3o2bo3bo2b3o$2bobo4b
2ob3obob2o2b2o2bo$ob2obob3o3bobobobobob2o3bo$obobobob2ob2obo2b3obob4o$
o2bob10o2b2o2b3obobob2o$3o2bo2b2o3bob2o2bobob2o$ob3o2b2o7b4ob3obo2bo$
3bob2obob2o4b2obo2b3o4b2o$b3o3bo3bo2b2o2bob2obobo3b2o$b3o4b2ob2o2bo9bo
bobo$b4obo4b2obo6b2obo4bo$2b2ob2obo2bobo4b2o4b2o2bo$b2o7b4o3b2obo2b2o
3b3o$2obob7o3bo2b3ob2o3bob2o$4b3ob2obo3b3o3b3o3b2o$obo4b2ob2ob4o3bob4o
2bobo$o3b4o2bob3obo6bobob4o$b4o2bobobob4ob2obo5bob2o$2b2o3b2o2bo5bobob
4ob2o2bo$o2bo5b3ob2o3b2o3bo2b3o!

[wwei23]

Code: Select all

RLE of a wick

WOAH, THAT WICK IS AWESOME!!!

[bubblegum]

Probably, although the percentage of yl384s in C1 and G1 that have 1 occurrence from a soup beginning with boooo is apparently 0, so I can't verify.

Try zz_LINEAR or PATHOLOGICAL. The debris at the back would prevent this being marked as a yl384.

Thanks.
It's a zz_LINEAR, by rliston.

[dvgrn]

It's a zz_LINEAR, by rliston.

Linkssss! We wantssss linkssss!

Specifically it's the ninth zz_LINEAR.

[MikeP]

Another strange one:

Code: Select all

#CXRLE Pos=-58,-60 Gen=18609
x = 116, y = 124, rule = B3/S23:T116,132
2o27b2o27b2o27b2o$2bo25bo2bo25bo2bo25bo2bo25bo$bo26bobo26bobo26bobo26b
o$o28bo28bo28bo$9b2o27b2o27b2o27b2o$8bo2bo25bo2bo25bo2bo25bo2bo$9b2o
27b2o27b2o27b2o2$9bo28bo28bo28bo$8b3o4bo8bo12b3o4bo8bo12b3o4bo8bo12b3o
4bo8bo$7bo3bob2obo6bo2bo9bo3bob2obo6bo2bo9bo3bob2obo6bo2bo9bo3bob2obo
6bo2bo$o5bo2b4obobo7bob4o5bo2b4obobo7bob4o5bo2b4obobo7bob4o5bo2b4obobo
7bob3o$6bo2bo2bo2bo8bobo8bo2bo2bo2bo8bobo8bo2bo2bo2bo8bobo8bo2bo2bo2bo
8bobo$5bobobob2o14bo6bobobob2o14bo6bobobob2o14bo6bobobob2o14bo$b2ob3ob
2ob2o15bob2ob3ob2ob2o15bob2ob3ob2ob2o15bob2ob3ob2ob2o15bo$b3obo3b2o19b
3obo3b2o19b3obo3b2o19b3obo3b2o$5bo3bo24bo3bo24bo3bo24bo3bo$3b3o2b2o22b
3o2b2o22b3o2b2o22b3o2b2o$6bob2o25bob2o25bob2o25bob2o$2bobo4bo21bobo4bo
21bobo4bo21bobo4bo$bo2b2obo22bo2b2obo22bo2b2obo22bo2b2obo$6bobo26bobo
26bobo26bobo$b5o24b5o24b5o24b5o$3bo3b2o23bo3b2o23bo3b2o23bo3b2o$5bobo
26bobo26bobo26bobo9$2o27b2o27b2o27b2o$2bo25bo2bo25bo2bo25bo2bo25bo$bo
26bobo26bobo26bobo26bo$o28bo28bo28bo$9b2o27b2o27b2o27b2o$8bo2bo25bo2bo
25bo2bo25bo2bo$9b2o27b2o27b2o27b2o2$9bo28bo28bo28bo$8b3o4bo8bo12b3o4bo
8bo12b3o4bo8bo12b3o4bo8bo$7bo3bob2obo6bo2bo9bo3bob2obo6bo2bo9bo3bob2ob
o6bo2bo9bo3bob2obo6bo2bo$o5bo2b4obobo7bob4o5bo2b4obobo7bob4o5bo2b4obob
o7bob4o5bo2b4obobo7bob3o$6bo2bo2bo2bo8bobo8bo2bo2bo2bo8bobo8bo2bo2bo2b
o8bobo8bo2bo2bo2bo8bobo$5bobobob2o14bo6bobobob2o14bo6bobobob2o14bo6bob
obob2o14bo$b2ob3ob2ob2o15bob2ob3ob2ob2o15bob2ob3ob2ob2o15bob2ob3ob2ob
2o15bo$b3obo3b2o19b3obo3b2o19b3obo3b2o19b3obo3b2o$5bo3bo24bo3bo24bo3bo
24bo3bo$3b3o2b2o22b3o2b2o22b3o2b2o22b3o2b2o$6bob2o25bob2o25bob2o25bob
2o$2bobo4bo21bobo4bo21bobo4bo21bobo4bo$bo2b2obo22bo2b2obo22bo2b2obo22b
o2b2obo$6bobo26bobo26bobo26bobo$b5o24b5o24b5o24b5o$3bo3b2o23bo3b2o23bo
3b2o23bo3b2o$5bobo26bobo26bobo26bobo9$2o27b2o27b2o27b2o$2bo25bo2bo25bo
2bo25bo2bo25bo$bo26bobo26bobo26bobo26bo$o28bo28bo28bo$9b2o27b2o27b2o
27b2o$8bo2bo25bo2bo25bo2bo25bo2bo$9b2o27b2o27b2o27b2o2$9bo28bo28bo28bo
$8b3o4bo8bo12b3o4bo8bo12b3o4bo8bo12b3o4bo8bo$7bo3bob2obo6bo2bo9bo3bob
2obo6bo2bo9bo3bob2obo6bo2bo9bo3bob2obo6bo2bo$o5bo2b4obobo7bob4o5bo2b4o
bobo7bob4o5bo2b4obobo7bob4o5bo2b4obobo7bob3o$6bo2bo2bo2bo8bobo8bo2bo2b
o2bo8bobo8bo2bo2bo2bo8bobo8bo2bo2bo2bo8bobo$5bobobob2o14bo6bobobob2o
14bo6bobobob2o14bo6bobobob2o14bo$b2ob3ob2ob2o15bob2ob3ob2ob2o15bob2ob
3ob2ob2o15bob2ob3ob2ob2o15bo$b3obo3b2o19b3obo3b2o19b3obo3b2o19b3obo3b
2o$5bo3bo24bo3bo24bo3bo24bo3bo$3b3o2b2o22b3o2b2o22b3o2b2o22b3o2b2o$6bo
b2o25bob2o25bob2o25bob2o$2bobo4bo21bobo4bo21bobo4bo21bobo4bo$bo2b2obo
22bo2b2obo22bo2b2obo22bo2b2obo$6bobo26bobo26bobo26bobo$b5o24b5o24b5o
24b5o$3bo3b2o23bo3b2o23bo3b2o23bo3b2o$5bobo26bobo26bobo26bobo9$2o27b2o
27b2o27b2o$2bo25bo2bo25bo2bo25bo2bo25bo$bo26bobo26bobo26bobo26bo$o28bo
28bo28bo$9b2o27b2o27b2o27b2o$8bo2bo25bo2bo25bo2bo25bo2bo$9b2o27b2o27b
2o27b2o2$9bo28bo28bo28bo$8b3o4bo8bo12b3o4bo8bo12b3o4bo8bo12b3o4bo8bo$
7bo3bob2obo6bo2bo9bo3bob2obo6bo2bo9bo3bob2obo6bo2bo9bo3bob2obo6bo2bo$o
5bo2b4obobo7bob4o5bo2b4obobo7bob4o5bo2b4obobo7bob4o5bo2b4obobo7bob3o$
6bo2bo2bo2bo8bobo8bo2bo2bo2bo8bobo8bo2bo2bo2bo8bobo8bo2bo2bo2bo8bobo$
5bobobob2o14bo6bobobob2o14bo6bobobob2o14bo6bobobob2o14bo$b2ob3ob2ob2o
15bob2ob3ob2ob2o15bob2ob3ob2ob2o15bob2ob3ob2ob2o15bo$b3obo3b2o19b3obo
3b2o19b3obo3b2o19b3obo3b2o$5bo3bo24bo3bo24bo3bo24bo3bo$3b3o2b2o22b3o2b
2o22b3o2b2o22b3o2b2o$6bob2o25bob2o25bob2o25bob2o$2bobo4bo21bobo4bo21bo
bo4bo21bobo4bo$bo2b2obo22bo2b2obo22bo2b2obo22bo2b2obo$6bobo26bobo26bob
o26bobo$b5o24b5o24b5o24b5o$3bo3b2o23bo3b2o23bo3b2o23bo3b2o$5bobo26bobo
26bobo26bobo!

From this soup:

Code: Select all

#CXRLE Pos=-14,-16
x = 29, y = 33, rule = B3/S23:T29,33
obobo2b2o2b2o4b5o2bob2o$4b3o2bobo6b7o3bo$2b2o3b5o3bobobo2bobobobo$ob3o
b2o2bo3b2o6b2o2bobo$ob2o2b3ob3ob2obo3b2ob2o$2bo2b2o2b2ob4obo4bob4o$ob
2ob2o3b3o2bo3bobo6bo$2obob5ob2o4b2ob3o2bo$o2bo4b2o2bo2bo7b2obobo$b2ob
2o4bobob12obo$bo4bo2bobo3bob3o2b3o2bo$3bo4b4obob4o4b6o$2b2ob4ob3obo4bo
4b2o$o2bo3b2o4b4o4bo2b5o$bobo3b2obo2bobob2ob3obob2o$2o3bobo3bo2bob2obo
b2o3b2o$ob2o2bob2o2b3o2bobo2b2ob2o$3o4b3o3bobob2o2bo5bo$bob4o2bobo2b2o
bo4bo3bobo$6ob2ob3o4b4o2b6o$b2ob2obob3obo10b2o2bo$o2bo3b4o4bob2ob2o2b
3o$b2ob2obobo2bobobo2b2o3bo$2o2b2o2b4obo2b3o2b2o2b3o$3o2b4obo7bo2bo2b
2obo$b2ob2ob2ob2ob3ob3o3bo4bo$b2obobo3b8ob3ob2ob2o$bo6b4ob4o2b3o2bo2bo
$4o4bob3o3bo7b2o$2b3o2bobob2o2b4ob2obo2bobo$4bo2bob2o5b2obob7o$obob2ob
2ob2o2bobob3ob2obo$b4obo4bo2bob2o2b4o2b2o!

I started this round of searching on a whim, purely as a way to try and hot-rod this little FPGA dev board that I have, but it's ended up finding curiosities quickly enough to keep me engaged.

Maybe I can find an agar that dissolves into gliders, or mostly gliders, at some phase of its cycle, that could form the core of a new engineered spaceship.

I have a slightly larger FPGA board, but repurposing old Bitcoin mining kit might get me more bang for the buck ... hmm.

EDIT: Well, what do you know, as I was writing up this post, the searcher found a second soup that yields the same agar. Since the first one was found after about 500 million soups, the odds against this seem astronomical. Maybe my RNG isn't as random as I'd like.

Code: Select all

#CXRLE Pos=-14,-16
x = 29, y = 33, rule = B3/S23:T29,33
o2b3o4bobo3bo3b2ob2ob2o$2ob5ob5obo3b2obob4o$ob2obo2b7obo3b5o$b2o3bob4o
b2o3b2ob2o2b4o$ob4obo2b2obobob2obobo4bo$2obo2b5ob3o2bob4ob2o$o3b3obob
3o3bo5b4o$bo5b2ob2o2b2o2bo2b2o2b4o$bob4ob2ob3o2bo2b2ob5obo$3o2b3o3bob
4ob3obo2bo2bo$o4bob3o3bo2b2ob2obobo3bo$ob3o3bo2b2ob4o2bo2b2o2bo$2b4o3b
ob2o3b4o2bobo2bo$2bo2bo5b3o2bobobob3o2bo$b2obo2bob3o3bobo6bobo$b2obo2b
6o4b2ob4o2b2o$5obob3o2b2ob3o3b3ob2o$bo2b3o3bob2o3bobo3b2o3bo$3o2bo2b2o
6b2ob2o2b4obo$o6bobob5o2b2o2b2o3bo$5obo3bo5bo2bobo4b2o$b2ob3ob7obo2b2o
4bob2o$ob3obob3o2b3ob2o2b2o4b2o$bo2b3obo4b2obob2obo4b3o$2obo2b5o3bobob
2ob2o3b2o$2bob3obob2o2bo3bobo4b2o$2o2bob2o2b2o3b2ob2o3b2obobo$3b2o2b2o
bob3ob2ob5o3bo$bob3o4bobob2o4bobobo$2b5obo2bobo2bobobo3bo2bo$3bobob2o
2bo4bob3o3b2o$5ob3ob2obob3o2bo2bo2b2o$b3o3bo2bo2b2obob2o2bob4o!

[dvgrn]

EDIT: Well, what do you know, as I was writing up this post, the searcher found a second soup that yields the same agar. Since the first one was found after about 500 million soups, the odds against this seem astronomical. Maybe my RNG isn't as random as I'd like.

Maybe not _that_ astronomical. Run it for a few months and I bet you'll see this one crop up a few times more. It has a few not-too-improbable bottleneck points, like

Code: Select all

x = 6, y = 10, rule = B3/S23:T29,33
4bo$3bobo$3bobo$4bo2$2b2o$bo2bo$o$o3b2o$5o!

The two source soups don't seem to be suspiciously similar to each other, so I kind of doubt your random-number generator has much to do with it.

[MikeP]

I have four seeds for this agar now, out of about 1.461 billion candidates tested, making its frequency about 2.7 parts per billion which seems astonishingly high; the patterns I posted earlier in the thread appeared after running searches for much longer, and I only have a single soup for each.

I doubt I'll leave it running on T29,33 for months, but I'd like to see how many more times this crops up over the next day or so.

[MathAndCode]

Maybe I can find an agar that dissolves into gliders, or mostly gliders, at some phase of its cycle, that could form the core of a new engineered spaceship.

How do you plan to overcome the problem that the agar emitting gliders will likely depend on reacting with its neighbors (or with itself in a toroidal universe)? Unless it's only one tile thick, which would prevent these likely crucial interactions from occurring, the agar can't simply be constructed by gliders emitted by previous copies of the same agar because some of the gliders would collide too early.
Keep in mind that I still consider myself fairly new here, so I could be wrong about this, or maybe I'm misunderstanding why you want an agar that's mostly gliders at some point.

[mniemiec]

EDIT: Well, what do you know, as I was writing up this post, the searcher found a second soup that yields the same agar. Since the first one was found after about 500 million soups, the odds against this seem astronomical. Maybe my RNG isn't as random as I'd like.

Maybe not _that_ astronomical. Run it for a few months and I bet you'll see this one crop up a few times more. It has a few not-too-improbable bottleneck points, like ...

Even easier: if you advance that 8 generations, it turns into a collision of an R pentomino and a forming honeyfarm - both of which are very common.

[LunaticPhantasm]

Code: Select all

x = 16, y = 16, rule = B3/S23
obbbbobooobbbboo$
bobobboooobooobb$
bbbbbobboboobooo$
obbooooobbbbbbbo$
bobboobbbobbbobo$
ooobbbbbobbobbbb$
ooobbbobboobbooo$
booooobboboooooo$
bbbooobobbooobob$
obboobobbboooobb$
booooooobooobboo$
obboobbobbbobooo$
oobboooobbobobob$
bobbboboboobbooo$
boobboobbbobooob$
bboboobbbbobbbbb!

First HWSS on HWSS 10 in C1, twelfth overall.

xq4_27deee6zwsighhga

The haul

[MathAndCode]

Code: Select all

x = 16, y = 16, rule = B3/S23
obbbbobooobbbboo$
bobobboooobooobb$
bbbbbobboboobooo$
obbooooobbbbbbbo$
bobboobbbobbbobo$
ooobbbbbobbobbbb$
ooobbbobboobbooo$
booooobboboooooo$
bbbooobobbooobob$
obboobobbboooobb$
booooooobooobboo$
obboobbobbbobooo$
oobboooobbobobob$
bobbboboboobbooo$
boobboobbbobooob$
bboboobbbbobbbbb!

First HWSS on HWSS 10 in C1, twelfth overall.

xq4_27deee6zwsighhga

The haul

The first HWSS saved the second HWSS. Otherwise, this would have happened.

Code: Select all

x = 4, y = 7, rule = B3/S23
b3o$o2bo$3bo$3bo$3bo$3bo$bo!

However, the domino spark suppressed the deadly banana spark.
A LWSS or MWSS missing one cell at the back in the same way would survive without help.