Post 9ki4MhrpJ8eabfGd84 by feonixrift@hackers.town
 (DIR) More posts by feonixrift@hackers.town
 (DIR) Post #9kgAK3HARJ0urp6rz6 by cwebber@octodon.social
       2019-07-09T15:03:06Z
       
       1 likes, 2 repeats
       
       Racket is an acceptable Python https://dustycloud.org/blog/racket-is-an-acceptable-python/
       
 (DIR) Post #9kgAK3W3XwvJc04lBw by Angle@anticapitalist.party
       2019-07-09T15:17:25Z
       
       0 likes, 0 repeats
       
       @cwebber Huh, I need to take a closer look at racket then. How is it in terms of computational efficiency? If I want to screw around with some relatively heavy duty computing, how does it compare? :/
       
 (DIR) Post #9kgAK47dIEytUXellI by cwebber@octodon.social
       2019-07-09T15:21:20Z
       
       0 likes, 0 repeats
       
       @Angle It's reasonably good.  https://ecraven.github.io/r7rs-benchmarks/
       
 (DIR) Post #9kgAK4OeGyamLJcMHg by Angle@anticapitalist.party
       2019-07-09T15:47:35Z
       
       0 likes, 0 repeats
       
       @cwebber Huh, interesting. Whats with the performance on ctak, fibc, and sum1? And how important are those? I'm not super experienced with optimization, but they stood out because of how oddly poorly racket scores on them, compared to everything else. I also note that sum1 says it's primarily a test of floating point arithmetic. If I happen to be working with neural nets, will that cause problems for me? Or, am I overthinking things? XD
       
 (DIR) Post #9kgAK4ahYAEWwhFz4S by cwebber@octodon.social
       2019-07-09T15:53:05Z
       
       0 likes, 0 repeats
       
       @Angle Some people here are using racket for machine learning.  Can't say I know anything about it myself.
       
 (DIR) Post #9kgAK4zs2aMmCl25Im by cwebber@octodon.social
       2019-07-09T15:53:29Z
       
       0 likes, 0 repeats
       
       @Angle But probably if you're doing neural nets you may be calling out to opencl? ;)
       
 (DIR) Post #9kgAK5QoQPuvYJdbIO by Angle@anticapitalist.party
       2019-07-09T15:54:59Z
       
       0 likes, 0 repeats
       
       @cwebber Well, I'm still screwing around in python right now. But, eventually, maybe. :/
       
 (DIR) Post #9kgAK7U4nFiTv1BEf2 by cwebber@octodon.social
       2019-07-09T15:59:16Z
       
       2 likes, 0 repeats
       
       @Angle I will say that in pretty much every category, Racket is faster than Python.The reason Python's neural net tools are fast is they don't use Python to do the heavy lifting. ;)
       
 (DIR) Post #9ki4MhXyUwm3c5ymBc by _emacsomancer@linuxrocks.online
       2019-07-10T13:19:06Z
       
       1 likes, 0 repeats
       
       @cwebber I think other than initially looking a bit funny or being unfamiliar to people who learned programming with non-Lispy languages, #Lisp #parentheses have more advantages than disadvantages. I mentioned to Matthew Flatt (re: paren-less Racket langs) that I think the parens could make Racket/Lisp potentially easier for getting #linguistics students into non-trivial programming than other langs, since linguists are used to dealing with #trees and bracket-notation for trees.
       
 (DIR) Post #9ki4MhrpJ8eabfGd84 by feonixrift@hackers.town
       2019-07-10T14:05:40Z
       
       0 likes, 0 repeats
       
       @_emacsomancer @cwebber and rainbow highlighting does more for readability than brackets do.
       
 (DIR) Post #9ki4Mi8UJBytRL3w6C by cwebber@octodon.social
       2019-07-10T14:12:10Z
       
       0 likes, 0 repeats
       
       @feonixrift @_emacsomancer Yes, I think this is a key thing often missed: reading parenthetical syntax can be *more readable* than non-parenthetical syntax by a dramatic amount with rainbow parentheses https://dustycloud.org/tmp/emacs_lisp_setup.pngAnd that's not even including how parenthetical structure makes code flow more obvious.
       
 (DIR) Post #9ki4MisvWX7zlMn1U0 by tija@pleroma.uwah.moe
       2019-07-10T14:45:18.783785Z
       
       0 likes, 0 repeats
       
       @cwebber @_emacsomancer @feonixrift > readable ­578598816_303826.gif
       
 (DIR) Post #9ki55PmhNxcrYUBfzE by epicmorphism@satania.space
       2019-07-10T14:53:26.539954Z
       
       0 likes, 0 repeats
       
       @cwebber @_emacsomancer @feonixrift the fact that you can use square brackets also helps a lot!<-- every time i write a let binding i have to think of the number of opening parens... using square brackets can make it muchmore enjoyable
       
 (DIR) Post #9ki58t0dBXWH3xML3I by epicmorphism@satania.space
       2019-07-10T14:54:04.625985Z
       
       0 likes, 0 repeats
       
       @cwebber @_emacsomancer @feonixrift also, are you writing python in that screenshot? :SatanaThumbsUp:
       
 (DIR) Post #9ki7HSgecTihwNES8W by cwebber@octodon.social
       2019-07-10T15:08:11Z
       
       1 likes, 0 repeats
       
       @epicmorphism @feonixrift @_emacsomancer Hy, so yes :)
       
 (DIR) Post #9kiEGvdiBl1N7nY5Ee by technomancy@icosahedron.website
       2019-07-10T15:27:31Z
       
       0 likes, 0 repeats
       
       @cwebber @feonixrift @_emacsomancer I'm going to be contrarian here and say that while yes, rainbow parens make it easier to keep your parens balanced, this isn't actually a good thing because balancing parens is a job for computers, not for humans.I have my theme set up so the parens fade into the background and are barely distinguishable; I find it more readable to not be distracted from the indentation.
       
 (DIR) Post #9kiEGvsxH5DLt4gFzk by technomancy@icosahedron.website
       2019-07-10T15:28:43Z
       
       1 likes, 0 repeats
       
       @cwebber @feonixrift @_emacsomancer (on the other hand, rainbows can be good for pair programming because you can easily indicate where in the code you mean--"what if we delete the green form?" which comes in handy, but for completely different reasons)
       
 (DIR) Post #9kiET9tB2FmYnGc8Zs by vk@mamot.fr
       2019-07-09T16:36:43Z
       
       0 likes, 0 repeats
       
       @cwebber It's so sad most useful lisps have the same bytes/strings duality problems as python :(
       
 (DIR) Post #9kiETAoFc3R5eHJiXQ by cwebber@octodon.social
       2019-07-09T16:41:29Z
       
       0 likes, 0 repeats
       
       @vk Racket doesn't?  (Also Python hasn't since 3.0?)
       
 (DIR) Post #9kiETB72UCSsaY6ip6 by vk@mamot.fr
       2019-07-09T16:51:25Z
       
       0 likes, 0 repeats
       
       @cwebber Unfortunately, I think all schemes have it (duplicate interface for string functions and file streams, decoding errors, and time / copies spent decoding) because it is required by r6rs that all "strings" and #\characters should be decoded (http://www.r6rs.org/final/html/r6rs/r6rs-Z-H-14.html#node_sec_11.11).For python it is a little more complicated but since version 3 it has similar characteristics.
       
 (DIR) Post #9kiETBTN9AKThoYYdM by cwebber@octodon.social
       2019-07-09T16:53:03Z
       
       1 likes, 0 repeats
       
       @vk oh, I misunderstood.  Duality of that is a feature, IMO.  Way less unicode errors if you deal with that stuff up front.
       
 (DIR) Post #9kiEYQCxmxJSr2NXg8 by gnomon@mastodon.social
       2019-07-09T16:47:39Z
       
       1 likes, 0 repeats
       
       @cwebber "excuse me, this snek is just six thousand parens in a trenchcoat"
       
 (DIR) Post #9kiEfawMHjRSAP244u by pjotrp@mastodon.technology
       2019-07-10T14:10:14Z
       
       1 likes, 0 repeats
       
       @cwebber I'll add that for me Racket makes programming fun again. Going from Ruby to Python was painful. Now I feel liberated.
       
 (DIR) Post #9kmItfXBB8Tqux3r8a by clacke@libranet.de
       2019-07-12T15:46:53Z
       
       0 likes, 0 repeats
       
       @technomancy @_emacsomancer @cwebber @feonixrift I hate it when machines help me and insert parenthesis for me and things like that. That said, I've never used a real structured editor, only text editors that pretend like they could be one.