Some info on L-systems for procedural generation in No Man’s Sky: http://nomanssky.wikia.com/wiki/Procedural_generation

Maybe not exactly L-systems, but close, for procedural building modeling: http://citeseer.ist.psu.edu/viewdoc/summary;jsessionid=3385B7D0876F865A2290595DED58F582?doi=10.1.1.112.183

Are you working on a related work section or just sending a pile or links that someone else would have to sort out later?

That is, should I be keeping notes here?

For now, just sending out links.

I’m chasing pointers a bit, trying to find interesting earlier papers, but not writing a section.

And some of that may be more intro material.

Is that useful, or should I stop?

The intro is about racket and building languages, I expect.

Or L-system motivation, wherever that would go.

It just reads like you are adding something to someone’s todo list.

But without saying whose.

I think digging into the related work on l-systems is helpful.

Adding thing to an unspecified person’s todo list is less helpful :slightly_smiling_face:

Oh, I didn’t see it as todo list items, but rather as a first pass of finding related work.

ok

I found a few more that I can post, but I can’t do much more than that first pass right now. Will have to do some errands and chores soon.

So if you think it may be useful, I can post them.

sorry I spoke. please do what you think is best.

(teh last 20 or so hours have been very slow and I’m frustrated, but not with you)

No worries. I think I see what you meant. Will post what I have.

City (well, road-grid, mostly) generation: https://graphics.ethz.ch/Downloads/Publications/Papers/2001/p_Par01.pdf

(“Official”, from ACM DL, for meta-info: http://dl.acm.org/citation.cfm?id=383292 )

Ok, that’s all the interesting ones I found so far.

Lots of uninteresting stuff, though.

This might be weird, but why would putting multiple l-system pieces cause “identifier used out of context” error?

This piece of code runs just fine: `(,#'module mod racket/base (,#'module post-hyphens racket (provide (all-defined-out)) (define (start variables) (cons 0 0)) (define (finish pr variables) (* 1. (if (zero? (cdr pr)) +inf.0 (/ (car pr) (cdr pr))))) (define (A pr variables) (cons (+ (car pr) 1) (cdr pr))) (define (B pr variables) (cons (car pr) (+ (cdr pr) 1)))) (require (prefix-in ::: (submod "." post-hyphens)) lindenmayer/lang) (l-system :::start :::finish '#hash((n . 5)) (A) (A -> A B) (B -> A))) But If I put another copy of l-system, the error shows up

@robby I want to push a file that has a collection of links for related work so we don’t need to look through slack history. Should that go in the repo you sent us, or is there another place for the paper itself?

that would be great

I plan to put the paper in that repo too.

It probably would work to put the pdfs of the papers into the repo (in their own directory) too.

on it!

Okay, I’ve pushed a small start to related work. I’ve having trouble finding a PDF for two of the papers, might have to actually go and ask the library or something.

Thanks. I’ve added abop to that.

And finally written some text.

It explains the runtime representation. Does it make sense?

@shu—hung sorry I missed your comment until now!

I am not sure. I guess I need more context to figure it out.

@robby looks good. One suggestion: It might be good in section 2 to mention that the Axiom is the starting string.

okay.

I’ve pushed an edit

@florence did you find the seminal cite?

The book seems to suggest it is “Mathematical Models for Cellular Interactions in Development ”

for l-systems?

but flipping thru it isn’t conclusive.

(yes)

OH, i think so, i just didn’t add in

whoops

@robby pushed

(yes, that is the origional citation I found as well, tho its odd, b/c the paper is in two parts, both published in the same issue of theoretical biology)

I didn’t see anything I recognized as an l-system in the first one.

But I didn’t spend much time on it yet

@robby the citation train I’m following claims this paper (http://www.sciencedirect.com/science/article/pii/0022519371900026) reformulates the l-systems into what we see now. What’s in this second paper seems to be what see as l-systems.

Looking at paper 2, what’s in paper 1 might be the basic of parametric l-systems, but I’m not entirely clear on what’s going on yet.

thanks for sorting this out.

I’m also not quite sure yet where context-sensitive l-systems enter the picture

We won’t talk about them in the paper, except perhaps to point out that the implementation strategy we use won’t work for them :stuck_out_tongue_winking_eye:

Okay, Ive think ive figured in out. There are three kinds of l-systems: 0L, 1L, and 2L. 0L is what we have been using (technically DOL-systems, but whatever). 1L and 2L and similar, but the productions the symbol the right of them (in 1L) and symbol to the left of them (2L) as extra parameters, and can dispatch on them. What’s introduced in the original paper is a 1L system. The second paper I linked to introduces OL systems. 2L systems are the context sensitive ones that we know. (Interpolating a little from the two papers and section 1.8 of the book for that one). 2L systems are introduces somewhere between the two papers, and I can track down where if we want.

(Technically, according to the book, its really IL systems for any I you want, but that just makes my head hurt)

also @robby should this channel be private?

yes.

how do I make it so?

The web says “gear icon -> additional options"

I see I could have picked that when I created it.

but only from the web?

nope

oh

oh “By default, Team Owners and Admins can use this feature. ”

argh.

isn’t software grand?

Do we need anything in the history here?

I’ve created lindenmayer.

We have vincent’s links in the .txt file, right?

Yes,

okay, I’ll delete this channel.

lets move