jonathanhwchan
2020-7-28 08:01:18

@jonathanhwchan has joined the channel


spdegabrielle
2020-7-28 15:31:59

Have I missed any tools that should be highlighted for standard-fish competition?


gknauth
2020-7-28 16:02:58

I’m giving a talk on celestial navigation on Friday and I’m curious to find out whether slideslow + metapict will do everthing I need. http://soegaard.github.io/docs/metapict/metapict.html


gknauth
2020-7-28 16:04:11

I guess the biggest question is how I’ll draw some of the 3D visualizations, e.g., great circles and circles of position on a sphere.


laurent.orseau
2020-7-28 16:20:26

@soegaard2


soegaard2
2020-7-28 16:21:53

I haven’t tried drawing spheres - but there is support for circles and ellipses.


soegaard2
2020-7-28 16:22:24

spdegabrielle
2020-7-28 16:24:31

where is your utah teapot example?


soegaard2
2020-7-28 16:24:55

soegaard2
2020-7-28 16:25:07

soegaard2
2020-7-28 16:25:26

The teapot was just a quick hack. I think you listed it under “artifacts”.


spdegabrielle
2020-7-28 16:25:50

I think pict3d can be embedded in slideshow?


gknauth
2020-7-28 16:27:38

I’ll find out! Now I’m wondering if I’ll be able to do pict3d with controls (alter height of stars & circles of position).



spdegabrielle
2020-7-28 16:28:20

I put the pict3d version but metapict looks nicer


spdegabrielle
2020-7-28 16:28:35

butI cant find the metapict version :disappointed:



soegaard2
2020-7-28 16:42:20

An alternative is to use Asymptote (or Tikz). For Asymptote there is a rather good 3d library: http://asy.marris.fr/ressources/ASY_3D.pdf


gknauth
2020-7-28 16:46:00

@spdegabrielle That is totally wild, thank you! And @soegaard2 too, I’m just trying out yours as well. Thank you both!


spdegabrielle
2020-7-28 16:49:20

FWIW I think metapict will look better in a slideshow



wanpeebaw
2020-7-28 18:44:12

@samdphillips Hope there is a way to fix this problem. Efficiency is really matter.


samdphillips
2020-7-28 18:48:40

If it’s contracts then it becomes an exercise in which parts of the system can you relax the constraints.


samdphillips
2020-7-28 18:50:04

Or move the contracts out of the hot path.


samdphillips
2020-7-28 18:51:04

I’ve only done a quick profile, so I’m not 100% on what the fix is. Just from past experience.


samdphillips
2020-7-28 18:52:55

As a point of comparison using the continued-fraction package across the whole space took around 800s (~13 – 14 minutes)


laurent.orseau
2020-7-28 19:02:22

That’s a factor 3000 compared to my version :scream:


spdegabrielle
2020-7-28 21:48:58

@soegaard2 whats”flomat.rkt”? utah-teapot.rkt #lang racket/base (require racket/match racket/list racket/port racket/string "flomat.rkt")



frank
2020-7-29 01:46:17

@frank has joined the channel


joel
2020-7-29 03:21:02

Can someone offer a rough approach for how I would test a potential PR specifically for Scribble? Would I need to somehow uninstall the scribble package and then do raco pkg install --link from inside a local clone of the scribble repo?


sorawee
2020-7-29 03:29:01

Do raco pkg update --clone scribble inside a directory you want to host the Scribble clone


sorawee
2020-7-29 03:29:34

Then, you will get a copy of Scribble as a git repository, that Racket also uses it.


sorawee
2020-7-29 03:29:42

Note that it takes time to set things up


joel
2020-7-29 03:52:08

raco pkg update: package is not currently installed from a repository package: scribble current installation: (catalog scribble)


joel
2020-7-29 04:01:17

Ah, figured it out, raco pkg update --lookup --catalog <https://pkgs.racket-lang.org\|https://pkgs.racket-lang.org> --clone scribble


joel
2020-7-29 04:01:37

Thanks for putting me on course!


wanpeebaw
2020-7-29 05:18:13

@samdphillips Get ride of continued-fraction package and ~a string formatting brings 16x performance boost. But still way slower than @notjack’s version. Doing too much unnecessary computation. :laughing: