|
our programmers suck so lets have them do manual error handling with an underpowered type system oh and make them use lots and lots of threads
|
# ? Nov 17, 2018 16:53 |
|
|
# ? May 22, 2024 15:13 |
|
yeah its a stupid argument for sure. teaching someone to use the wrong tools because they're new to something is a terrible idea.
|
# ? Nov 17, 2018 17:06 |
|
generics are also rly easy to use imo
|
# ? Nov 17, 2018 21:35 |
|
rob pike's primary mistake is that he confuses the simplicity of implementing the language with the simplicity of using the language. all his arguments use the later to justify the former HappyHippo fucked around with this message at 23:02 on Nov 17, 2018 |
# ? Nov 17, 2018 23:00 |
|
all this being said, I always hear that actually writing Go is more enjoyable than its design flaws would imply. does that mean anything? who cares, Rust is fun as heck
|
# ? Nov 17, 2018 23:35 |
|
HappyHippo posted:rob pike's primary mistake is that he confuses the simplicity of implementing the language with the simplicity of using the language.
|
# ? Nov 18, 2018 00:32 |
|
Beamed posted:all this being said, I always hear that actually writing Go is more enjoyable than its design flaws would imply. does that mean anything? who cares, Rust is fun as heck
|
# ? Nov 18, 2018 02:26 |
|
gofmt is absolute genius it is the only good idea in all of golang
|
# ? Nov 18, 2018 03:00 |
|
I wonder how many people simultaneously decry IndentationError and laud gofmt
|
# ? Nov 18, 2018 03:03 |
|
pokeyman posted:I wonder how many people simultaneously decry IndentationError and laud gofmt the stupidity of IndentationError is that the python interpreter will accept tabs or spaces and heuristically guesses which it's looking at they didn't commit
|
# ? Nov 18, 2018 03:15 |
|
it’s not ideal but I probably would’ve made the same decision. if you want people to be able to use your language without loving around with their text editor, which is entirely plausible when one of your goals is to make an easily teachable language, you probably can’t pull a make and just fail when the other whitespace is encountered
|
# ? Nov 18, 2018 03:33 |
|
you already crossed that line when you designed a whitespace sensitive language
|
# ? Nov 18, 2018 03:52 |
|
regardless of the desirability of dealing with hosed up editor settings like using hard tabs it’s indefensible that the indentation heuristic is run per block so you can have a tab indented block containing a four space indented block containing a two space indented block
|
# ? Nov 18, 2018 04:44 |
|
the defense is that developers overwhelmingly don't know whether they're inserting tabs or spaces and can't be made to care, in that respect rob was right
|
# ? Nov 18, 2018 09:26 |
|
python's braindead approach to whitespace makes copying and pasting code from the internet very likely to completely mutilate it and require careful re-reading and re-indenting. as a way to deter novices from blindly kludging together code they haven't read, i almost like it almost
|
# ? Nov 18, 2018 17:15 |
|
ah yes making this language hard to use without enabling visual whitespace in my editor is a safety feature
|
# ? Nov 18, 2018 18:12 |
|
hey janitor, yesterday was forth day. you didn't miss much
|
# ? Nov 18, 2018 19:34 |
|
the guy who runs the only forth chips business in wyoming did a remote talk. they had been in debt for a while but some mystery benefactor in japan handed them enough bitcoin to put them in the black, without any formal strings, just the gentleman's agreement to make a new generation of chips. i hope the boss cashed that bitcoin on the spot, for the company's sake, but I can't be sure
|
# ? Nov 18, 2018 19:39 |
|
Gazpacho posted:the guy who runs the only forth chips business in wyoming did a remote talk. they had been in debt for a while but some mystery benefactor in japan handed them enough bitcoin to put them in the black, without any formal strings, just the gentleman's agreement to make a new generation of chips. i hope the boss cashed that bitcoin on the spot, for the company's sake, but I can't be sure cash injections from foreign mafias are not usually the sign of a thriving industrial enterprise
|
# ? Nov 18, 2018 19:45 |
|
Internet Janitor posted:python's braindead approach to whitespace makes copying and pasting code from the internet very likely to completely mutilate it and require careful re-reading and re-indenting. as a way to deter novices from blindly kludging together code they haven't read, i almost like it it does seem like a weird way to janitor where things go as opposed to using brackets on the other hand the decision was made way back so who knows
|
# ? Nov 18, 2018 20:09 |
|
Also this guy got several local farm owners to establish a nonprofit foundation the purpose of which is to staff his company with its students so he can teach them electronics and forth programming and eventually hire them.
|
# ? Nov 18, 2018 20:12 |
|
reasonable men can disagree over whether whitespace sensitivity was a good choice in python. but no sane person can defend heuristically guessing the indent level. they needed to make a single, correct, enforced-by-the-interpreter choice i don't even care which choice it was: 2 spaces, 4 spaces, hard tabs, I give zero shits. (every time i change teams i have to twiddle a setting in my editor and then i never look at it again.) the important part was that they needed to make a choice, and they didn't, so now it's half-baked confusing nonsense
|
# ? Nov 18, 2018 20:35 |
|
Guessing the indent type would be fine if it just used the first indentation it found through the whole file. I'm thankful it didn't end up like Makefiles
|
# ? Nov 18, 2018 20:42 |
|
All python did was integrate the status quo ante for people who actually indent their code, and that status was “indent each block some from its surroundings” with most programmmers not caring to specify or even know what the “some” is. When I had a horrible web app dropped on me at prev. job, one of several cleanup tasks was to replace the arbitrary mixture of tab/space indentation to just one spacing character. Then I handed off to the new overseas team who proceeded to mix it up again. Looks good in my editor, must be ok
|
# ? Nov 18, 2018 20:46 |
|
Boiled Water posted:on the other hand the decision was made way back so who knows the explicit predecessor of python was abc, which came with a syntax-directed repl. it's a bit interesting to try out. given syntax-directed editing, it makes some sense to minimize block delimiters, and thus syntactic noise. perhaps the design decision was taken forward to python without further consideration? one surprising aspect of abc is that there is no boolean datatype, only syntactic boolean expressions. you can define a predicate as a special case of function suitable for reference somewhere that syntactically needs a boolean, but you cannot store a boolean value in a variable: code:
|
# ? Nov 18, 2018 20:50 |
|
btw if you think whitespace heuristics in python are bad wait til you hear what apple does with swift code
|
# ? Nov 18, 2018 20:55 |
|
Notorious b.s.d. posted:but no sane person can defend heuristically guessing the indent level. they needed to make a single, correct, enforced-by-the-interpreter choice
|
# ? Nov 18, 2018 20:58 |
|
we live in a world where a computer forcing someone to do the only sane and unambiguous course of action can significantly hurt your language's adoption lol
|
# ? Nov 18, 2018 21:02 |
|
we live in a world where javascript
|
# ? Nov 18, 2018 21:12 |
|
RPATDO_LAMD posted:we live in a world where javascript a pro-tier post/av combination
|
# ? Nov 18, 2018 21:20 |
|
Gazpacho posted:btw if you think whitespace heuristics in python are bad wait til you hear what apple does with swift code i have never heard any actual complaints about how this works in practice, only “well if you read the spec it’s complicated” if you have a real problem with it you can always use semicolons
|
# ? Nov 18, 2018 21:27 |
|
swift has an open language/library evolution process and a lot of passionate users, so there are plenty of people participating who’ve never been part of a language effort before, or even really read a spec. there’s a certain subset of those people who are just repeatedly shocked that languages have, like, details, and because this is 2018 they like to get real mad about it on twitter
|
# ? Nov 18, 2018 21:49 |
|
"the rules are too complicated to commit to memory" is a legit complaint
|
# ? Nov 18, 2018 21:52 |
|
anecdotally from my own teaching experience, especially with younger kids and total beginners, it is quite difficult to get new students used to the semantic significance of whitespace. they had a somewhat easier time with languages that use explicit block delimiters. line-numbered BASIC- which notably does not feature any kind of recursive syntactic structure outside of arithmetic expressions- was by far the most intuitive syntax for these groups. students could easily form an accurate mental model of the step-by-step nature of a BASIC program. familiarity with this style of notation seemed to help later when introducing block structure by showing a correspondence to familiar patterns. if you already know how to program in some language, picking up python's syntax and semantics aren't very hard. i am, however, quite critical of its use as an introductory language.
|
# ? Nov 18, 2018 21:56 |
|
Gazpacho posted:"the rules are too complicated to commit to memory" is a legit complaint the rule is “don’t put the left-paren of a call on the next line, don’t put the left-bracket of a subscript on the next line, and don’t write binary operators right next to their operands so they look like prefix/postfix operators” no one ever does the first two so the only part you have to remember is the last
|
# ? Nov 18, 2018 22:02 |
|
swift is an extremely good language especially for beginners. it could and should replace Go
|
# ? Nov 18, 2018 22:20 |
|
semicolons good though
|
# ? Nov 18, 2018 23:49 |
|
hackbunny posted:swift is an extremely good language especially for beginners. it could and should replace Go same, but c#.
|
# ? Nov 18, 2018 23:54 |
|
redleader posted:semicolons good though
|
# ? Nov 19, 2018 00:14 |
|
|
# ? May 22, 2024 15:13 |
|
i've been writing nothing but swift for six months and I don't even know what the above discussion is referring to so it can't be that much of a hindrance. the piss poor error reporting is my only major complaint about it and at least I've developed a feel for intuiting what the actual problem is when I get nonsensical complaints about argument labels
|
# ? Nov 19, 2018 00:19 |