|
Cybernetic Vermin posted:the option to use a dedicated teaching language is there, but there'd be some real upsides to a language with some scientific computing, math, and visualization credentials as we funnel a lot of engineering disciplines through our intro courses. So teach them MATLAB
|
#
?
Feb 9, 2024 19:27
|
|
|
|
| # ? May 20, 2024 08:14 |
|
|
Plorkyeran posted:there was a time when the mainstream languages people wrote actual software in did not have closures and higher-order functions were arcane black magic, so lisp seemed incredibly fancy homoiconicity means lisp repls are still better and they can kramer in really serious language features with great backwards compatibility unlike the pythons of the world, but thats about it
|
|
#
?
Feb 9, 2024 19:38
|
|
|
Tunicate posted:So teach them MATLAB please dont
|
|
#
?
Feb 9, 2024 19:39
|
|
|
Tunicate posted:So teach them MATLAB one big motivation for me here is the complete elimination of matlab from our entire curriculum.
|
|
#
?
Feb 9, 2024 19:39
|
|
|
bob dobbs is dead posted:homoiconicity means lisp repls are still better and they can kramer in really serious language features with great backwards compatibility unlike the pythons of the world, but thats about it for sure
|
|
#
?
Feb 9, 2024 19:46
|
|
|
clojure core.async was shoved into the language as a first party dealio 8 years after language release and got serious uptake about 10 years after language release. no breaking changes
|
|
#
?
Feb 9, 2024 19:50
|
|
|
it is amazing how bad the default python repl is though. to a point where i imagine i could spend a day on improving it and if the patch was accepted it'd be the headline feature of the release for most normal users. but i imagine there's some politics or weird compatibility reason why it has to be so primitive
|
|
#
?
Feb 9, 2024 19:50
|
|
|
Internet Janitor posted:i think processing is excellent for teaching programming plus the ide is terrible, which sets expectations for programming appropriately
|
|
#
?
Feb 9, 2024 20:05
|
|
|
Cybernetic Vermin posted:it is amazing how bad the default python repl is though. to a point where i imagine i could spend a day on improving it and if the patch was accepted it'd be the headline feature of the release for most normal users. yea the ergonomics are trash if you gently caress up typing something or just wanna make a small change, you cant copypast the working parts because of the >>>/... so you basically either have to start over or arrow up and hit return each line again until the one you wanna edit, then hit return on the rest. which of course makes the history longer so you still have to arrow up 10 times for every line you basically have to edit your thing in a text file and then paste it into the repl. might as well just run the script directly with the interpreter Carthag Tuek fucked around with this message at 20:16 on Feb 9, 2024 |
|
#
?
Feb 9, 2024 20:14
|
|
|
i really like java as an intro language because it leads very naturally into the most important followup skill, modding minecraft
|
|
#
?
Feb 9, 2024 20:16
|
|
|
the way you do it w/ clojure nrepl is you just work in a text editor, select part of your program, because everything's a statement, you can select in text editor, gently caress w some part of the running program and rerun just that part of the program direct from the text editor and get the output in some other buffer the way you do it w/ common lisp, you can even avoid actually filling in statements and use the nutso common lisp error handler to fill in the statement in your text editor / in the program simultaneously as you run bob dobbs is dead fucked around with this message at 20:22 on Feb 9, 2024 |
|
#
?
Feb 9, 2024 20:19
|
|
|
i guess some stuff you can monkeypatch, but its not ideal
|
|
#
?
Feb 9, 2024 20:23
|
|
|
bob dobbs is dead posted:the way you do it w/ clojure nrepl is you just work in a text editor, select part of your program, because everything's a statement, you can select in text editor, gently caress w some part of the running program and rerun just that part of the program direct from the text editor and get the output in some other buffer never really got the hang of slime, but I could sense there was power within
|
|
#
?
Feb 9, 2024 20:28
|
|
|
bob dobbs is dead posted:just start recommending decker and be done with it lol in its current form i don't think decker would be very good for general introductory programming courses i have reference documentation, but an area i'm sorely lacking is a tutorial curriculum for scripting; i plan to work on this lil and the whole decker environment are very different from a conventional p-lang-and-text-editor workflow and these might inherently cause undue difficulty transitioning to java or python or whatever later i have had some encouraging discussions with educators about using it in intro game design or narrative design courses, though; the same areas where tools like twine are appropriate
|
|
#
?
Feb 9, 2024 20:41
|
|
|
one or our unis here uses racket for intro classes and everyone ive talked to that took that program despises it what makes me think it was the right choice
|
|
#
?
Feb 9, 2024 20:52
|
|
|
common lisp is fun Im just getting back into it and its nice being able to install any feature I might want as a library. wanted pattern matching and algebraic data types and its just one download away cant imagine working with it as a job for the same reasons
|
|
#
?
Feb 9, 2024 20:54
|
|
|
i learned to program in basic and i turned out just fine
|
|
#
?
Feb 9, 2024 21:01
|
|
|
no you didnt you post here like the rest of us
|
|
#
?
Feb 9, 2024 21:02
|
|
|
Internet Janitor posted:in its current form i don't think decker would be very good for general introductory programming courses oh man i would love for kids to learn to program using decker  i think i learned more about programming from hypercard than c64 basic, which doesnt mean much tbh, but now that i think about it, decker/hypercard is the perfect repl. youre just live editing your app the whole time. it owns bowns
|
|
#
?
Feb 9, 2024 21:02
|
|
|
bob dobbs is dead posted:no you didnt you post here like the rest of us Also I kind of think basic is probably even worse than c in terms of instilling bad habits (I guess depending on the version)
|
|
#
?
Feb 9, 2024 21:02
|
|
|
mystes posted:Also I kind of think basic is probably even worse than c in terms of instilling bad habits (I guess depending on the version) except the worst habit: being a loving elitist shithead about it while not being able to solve the most basic of computational tasks. the cs program here was never that strong, but i think the monoculture of insufferable (male, white, youth) people having strengthened over the last 20 years is to no small part attributable to a bizarre and irrelevant low-level focus, displayed (though certainly not consisting entirely) in the switch from sml to c as intro language.
|
|
#
?
Feb 9, 2024 21:12
|
|
|
c sort of makes sense as a stepping stone. like say starting with assembly language and observing that setting up stack frames and tracking which registers hold what is tedious, so let's move on to c. and then observing that manual memory allocation and a lack of reusable data structures is tedious, so let's move on to python or rust or whatever. ... well, except that this sort of bottom-up teaching is great for already-knowledgeable people to jerk off to and terrible for actually teaching newcomers because none of these zoomed-in minutiae have any bearing on any problem that they actually care about. by definition a novice does not know the big picture yet. probably better to start with something like python and then work your way downwards from there instead idk. depends on your teaching objectives i suppose? if you're teaching "data science" or whatever then python and jupyter etc are probably fine as both a beginning and end point.
|
|
#
?
Feb 9, 2024 21:26
|
|
|
bob dobbs is dead posted:no you didnt you post here like the rest of us i feel nervous about posting itt specifically to be honest with you because there are like, actual language designers in here
|
|
#
?
Feb 9, 2024 21:28
|
|
|
Sapozhnik posted:i feel nervous about posting itt specifically to be honest with you because there are like, actual language designers in here 
|
|
#
?
Feb 9, 2024 21:33
|
|
|
we language designers are a wild and fearsome race, long of tooth and sharp of claw. dark is the blood with which we anoint our specifications. our laughter is cruel, and our committees black havens of the depraved
|
|
#
?
Feb 9, 2024 22:14
|
|
|
rjmccall posted:we language designers are a wild and fearsome race, long of tooth and sharp of claw. dark is the blood with which we anoint our specifications. our laughter is cruel, and our committees black havens of the depraved unironically
|
|
#
?
Feb 9, 2024 22:15
|
|
|
do you do birthday parties?
|
|
#
?
Feb 9, 2024 22:15
|
|
|
not in less than five hours of discussion
|
|
#
?
Feb 9, 2024 22:17
|
|
|
Subjunctive posted:do you do birthday parties? nah, we can't agree on date representations
|
|
#
?
Feb 9, 2024 22:27
|
|
|
also why we’re slowly dying out
|
|
#
?
Feb 9, 2024 22:35
|
|
|
open issues - cake contents TBD [100 KPI] - candles per cake algorithm, calendrical? #TODO localize. might be 1-indexed - fix race condition in cut cake / first pick routine (maybe exclusive or?) - implement time skip for adults during package opening sequence - early return guard possible if kids have bad vibes? - bug tracker is at neverreturn@myhouse.local
|
|
#
?
Feb 9, 2024 22:43
|
|
|
Carthag Tuek posted:open issues party is over because someone brought a gift that was too big for the gift buffer
|
|
#
?
Feb 9, 2024 22:48
|
|
|
Armitag3 posted:party is over because someone brought a gift that was too big for the gift buffer it overflowed into the pool and I am NOT cleaning that up
|
|
#
?
Feb 9, 2024 22:55
|
|
|
Every time you think that to participate in language standardization, or writing the standard library or w/e requires you to be real smart, just remember: they let me do it.
|
|
#
?
Feb 10, 2024 01:41
|
|
|
stack canary died when someone cooked with a non stick pan
|
|
#
?
Feb 10, 2024 03:51
|
|
|
Internet Janitor posted:awt and swing look dated and non-native, and are kind of a pain in the rear end to use, but they're functional. with a little effort you can get results comparable in many ways to Qt, but unlike Qt you don't have to run all your code through an insane custom preprocessor Counterpoint : put the stuff the moc enables in the language (some of it can be done natively now already). Then use Qt.
|
|
#
?
Feb 10, 2024 12:15
|
|
|
Tunicate posted:So teach them MATLAB MATLAB is a terrible language to teach programming fundamentals in but it might be really informative to have students do some sort of math heavy task in a more general purpose language, then compare how much easier it is to do the math parts in MATLAB? jk, I think anything that pedagogically innovative is probably a terrible idea in a large, “one size fits most, we hope” class
|
|
#
?
Feb 10, 2024 15:07
|
|
|
Just use numpy or something
|
|
#
?
Feb 10, 2024 15:15
|
|
|
Athas posted:Standard ML is an excellent teaching language. apropos of nothing, my alma mater outputs a ton of machine learning research and there was a tendency to refer to non-reenforcement learning techniques as "standard ml" which even as a budding plang thread poster 15 years ago never failed to throw me for a loop
|
|
#
?
Feb 10, 2024 15:35
|
|
|
|
| # ? May 20, 2024 08:14 |
|
|
more like p-nis
|
|
#
?
Feb 10, 2024 20:10
|
|
