@El'endiaStarman I'm actually following through with the blog idea.

\o/

In my opinion, my main criticism of the wireworld website is its lack of thoroughness.

It'll be nice to have one place where everything is pulled together.

Yeah, and I was thinking earlier that the learning curve is really steep. Like, from basics to ROM in seven steps.

I hope that, someday, this project becomes well-known within the GoL community.

...I should add my logo to VarLife sooner or later in anticipation of that... :P

but seriously though, this project is massive

definitely

It'd probably be a good idea to do something like primes before attempting Tetris.

Well, then again, since we'll have the assembly abstraction, I guess that's not all that necessary.

Is there any smaller size for GoL metapixels?

I do not believe so.

There are probably smaller metapixels, but not with the "all life-like rules" we've been assuming.

Well, the 'deep cell' is smaller, I think, but it's hard to tell if cells are alive or dead. The OTCA metapixel we're using was designed to be 2048 square to work well with HashLife, so that's basically the fastest useful metapixel we'll have.

Oh hey, I wonder if we could make a VoL metapixel!

Or, probably better, just figure out a pixel structure that can be turned on/off. Could probably use a series of pulses from the bottom/top and one of the sides that toggle or turn on a pixel when they collide.

I wonder how painful it would be to implement an RNG...

we do need something to randomize tetris pieces...

We can probably just use a low-quality randomization algorithm. It doesn't need to be cryptographically secure or anything...

PS, I'm calling myself the project leader in my blog post, because I can.

Something like a, a^2, a^3, ... % n where n is a large prime.

@PhiNotPi Well, you are the project leader...

I'm just the assistant. :P

@El'endiaStarman who all has made significant contributions to the project?

Conor and Mars Ultor, I think.

What did Mars do?

Uh, I'm not sure if they did anything. Checking the transcript now.

Well, they were named 'somebody' at that time, and they cloned VarLife and extended it a bit, but neither of us used it.

Both he and zyabin chatted a lot, and it looks like one or both of them did some stuff with Golly, which isn't a direct contribution to the project.

Yeah, I don't see any direct contributions from either somebody/Mars Ultor or Zyabin beyond discussions. You're project leader, I made the VarLife tool, and Conor will eventually make the circuitry tool.

Also, I verified that metapixels can be programmed independently and not break.

I actually did that a couple months ago.

why does wordpress think "Approach" is misspelt?

@CᴏɴᴏʀO'Bʀɪᴇɴ ^ don't make me regret this... :)

Posting here because this would benefit from the thoughts of people actively working on a solution

I'm writing a second blog post right now.

\o/

I'll be off soon to go to a Memorial Day parade, but I look forward to reading it when I get back.

Okay. This one is titled "intro to metapixels"

I'm kinda skipping over the details of how we go from GoL to metapixels, other than the fact that metapixels exist.

@Phi I don't think you'll regret it :3

@El'endiaStarman @CᴏɴᴏʀO'Bʀɪᴇɴ ^

I just read the first one. On that, I just got a new keyboard. Once I disable the original one, I should be good to go. (after an infection clears up a bit :P)

And now the second one. Its a nice clarification and a good read.

@CᴏɴᴏʀO'Bʀɪᴇɴ thanks!

looks like someone reverse-engineered the wireworld computer to give a more detailed explanation: logre.eu/wiki/Projet_Wireworld/en

@El'endiaStarman I'm working on the third blog post right now, about wires.

Awesome.

@El'endiaStarman Did you like my last two posts?

Yeah, they're good so far.

@El'endiaStarman @CᴏɴᴏʀO'Bʀɪᴇɴ ^

parallel wire crossing: goo.gl/y8ktAM

...isn't that the design I came up with?

@El'endiaStarman this was yours: goo.gl/cvLr3X

Exactly the same except for those two blue pixels.

yes, I know

Ohhh, those two blue pixels prevent double-tailed electrons.

yeah, I found a way to kill off the double-tails.

Don't worry, in my blog I put

> El’endia also gets most of the credit for this design

Good enough. ^_^

Have you read the third post yet?

I'm working through it.

> The corners listed above

@El'endiaStarman fixed, and also changed "failed attempt" to "almost-correct attempt", since I wrote the article originally before a working parallel cross was found.

Cool. Are you not going to include gifs of the delay wires?

@El'endiaStarman I could.... I think I was running into GIF issues

Were those issues due to VarLife?

@El'endiaStarman yeah, I'm trying again with fewer gens

Huh, I wonder if zyabin's problem was too many generations?

Have two electrons on the bottom wire.

(notice the skip, I would have to do 66 generations for a perfect loop, instead I did 33)

I think the skipping version is fine too.

Yeah, looks good to me.

> El'endia gets credit for this design

@El'endiaStarman ^ new wording

\o/

What I'm really waiting on right now is Conor's circuit simulator.

yeah

@CᴏɴᴏʀO'Bʀɪᴇɴ no pressure :)

>_>

working on it best I can :p

I think the next post will be about some basic gates, although I'm not sure what order I should put them it.

There's probably too many gates to fit in one post.

do basic gates, then comlicated ones

Don't we only have four kinds of gates? AND, OR, XOR, AND-NOT. You can put gifs of each configuration on one line, and if you're worried about size, you can resize the cells and the gifs will be sized accordingly.

hey @El'endiaStarman, I have the following:

let drag = d3.behavior.drag().on("drag", data => {
	let t = d3.select(this);
	t.attr("x", d3.event.x).attr("y", d3.event.y);
});
d3.selectAll("rect.tile").call(drag);

it doesn't work

@CᴏɴᴏʀO'Bʀɪᴇɴ I figured, since you're asking. :P

:p

t.attr("x", function(){return d3.event.x;}) and likewise for y.

o

but you use this code on your site:

 c.attr("cx", d3.event.x).attr("cy", d3.event.y);

Oh? Hmm.

Is t what it should be?

owait

it's probably because I'm using an anonymous fuanction

*arrow func

:D it works now

:D

next step is to implement a "snapping" behaviour

stands in corner

@CᴏɴᴏʀO'Bʀɪᴇɴ You can look at my Radians/Degrees demo for some inspiration on that.

wow, it's very convenient that some code exists! :D

:D

I'm not sure how much it'll help since my code is snapping to specific points within a tolerance, and yours should be snapping to a grid.

That's easy though, so [shrug].

true

General idea: look at current coors, find closest edge to it, and set disp x & y to that

What I would do is divide the current coors by some scale factor, round, then scale back up.

that sounds better :P

except idk what that scale factor would be

The size of the grid spacing?

o

like, the width of the cells?

yeah

so, since I have a padding factor, would I divide the coors by (cellWidth / paddingFactor)?

Padding factor...like, the spacing between cells? Just add that to the cellWidth.

well, I'm multiplying by the factor

tile => tile.x + padFactor * this.cellWidth * tile.x

Ah, okay. Then yes.

well it doesn't quite work :/

the function seems almost exponential

...wait, shouldn't you be dividing by (cellWidth * padFactor)?

ah that might be it

works beautifully

pain is not conducive to coding

heh :/

so close

there's, like, a pixel off

Yeah, there's usually a one-off kind of error that I just fix by smart trial-and-error.

ugh

It's probably because the padding factor isn't a nice binary number

oh well, I'll just fix it later

Yeah, it's a minor issue.

@CᴏɴᴏʀO'Bʀɪᴇɴ good work so far