2. This is not exactly a language feature but more of an attribute of the ecosystems, but I wish Common Lisp had a more comprehensive 'batteries included' standard library. It does have a pretty good standard library [1] for its time, but in the modern day, it fades in comparison to that of mainstream languages like Python or Go. Yes, the availability of Quicklisp compensates for the relatively smaller standard library, but it does not fully replace the value of having a large and cohesive standard library.
For Python, I wish (and this might be controversial) that it used s-expressions instead of whitespace-based delimiter for blocks. I find cutting, pasting, moving around, and in general, manipulating whole s-expressions as individual units very convenient with appropriate editor support (e.g. Paredit). I know there are languages like Racket, Clojure, etc. which fulfil the requirement of extensive batteries included library + s-expressions, but they are much less popular than Python. In my ideal world, one of the top three popular languages would be a Lisp or Lisp-like language with an extensive standard library comparable to Python's.
3. Common Lisp for personal projects. Python for collaborative projects.
1. Newspeak, Smalltalk
2. Smalltalk's code browser can be tedious, Newspeak's IDE is more coherent.
TCPSockets, tools
3. Smalltalk's Seaside framework for server side web, Newspeak for clients.
1. PHP and C.
2. I wish C had PHP's native "associative array as junk drawer" data structure. Conversely, I wish PHP had more support for a "long lifetime" task that wasn't expected to vanish at the end of a page load.
3. Am I trying to target the web? Then PHP.
2. There are no such features. But I want all languages to have J in the same manner as they have regexp. I mean a tool like NumPy but with sweeter syntax and more possibilities.
3. J is for number munching, Lisp is for everything else.
A version of Python (call it "Pith" [0]) that fixes the annoying things, eg:
no colons
if bool:
# should be
if bool
for thing in things:
# should be
for thing in things
nicer dicts
a_dict['prop']
# should be
a_dict.prop
shorter dicts
a_dict = {'name': 'value'}
# should be
a_dict = {name: 'value'} # like in js
# even
a_dict = {content-type: 'json'} # ooh, ahh !
auto main function
if __blah_blah_blah__ == '__main__' # dunder-struck ?
main()
# should be
# nothing -- it just calls main() if it's there
dict.get(prop) by default
# instead of
a_dict.get('prop')
# it should be like JS that returns undefined or None or something nullish or falsey
a_dict.prop
ternary
v = a if this else b # barf
# should be
v = this ? a : b
# or (to steal from VB)
v = iif(this, a, b)
# maybe fix list/dict-compros, too, while i'm at it
elif? elf? what the helf ?
elif x:
# should be
else if x # like any decent language
W T F-string?
f'{huh}'
# should be
`{huh}` # like JS
len (like it's BASIC)
len(a_string)
# sb
a_string.length
a longer example
# Regular Python | # Pith uses less punctuation
|
def main(): | def main
a_dict = { | a_dict =
'name': 'value' | name: 'value' # maybe use = instead of :
} |
if True: | if true
print(a_dict['name']) | print a_dict.name
elif 1 > 2: | else if 1 > 2
print('ooh') | print 'ooh'
|
main() | # no need to call main()
i also like how JS allows you to call a function before it's defined (but Python doesn't -- why !!!)
maybe add a do/while loop, multi-line comments ###, JS-style regex /reg/ instead of r"regex" blah blah blah
etc, etc, etc
it could be a preprocessor (like the C Preprocessor) that takes Pith and converts it to Python
for JS, take out parens, curly braces, semicolons, etc (make it "look" more like Pith)
i know these changes are fraught with peril, but i don't care (do i?). there's no reason (is there?) that i should be stuck with Guido's or Brendan's design choices [1] (i can make my own :) )
i even had an idea that you could write a program in Pith and it could output either Python or JS (or anything else). that might need a little more thought
[0] "Pith" is prolly already in use by something else, just humour me
[1] that's the bigger idea here -- take a language you like, fix all the things you don't like about it. maybe lisp with fewer parens
Answering my own questions:
1. Common Lisp, Python.
2. This is not exactly a language feature but more of an attribute of the ecosystems, but I wish Common Lisp had a more comprehensive 'batteries included' standard library. It does have a pretty good standard library [1] for its time, but in the modern day, it fades in comparison to that of mainstream languages like Python or Go. Yes, the availability of Quicklisp compensates for the relatively smaller standard library, but it does not fully replace the value of having a large and cohesive standard library.
For Python, I wish (and this might be controversial) that it used s-expressions instead of whitespace-based delimiter for blocks. I find cutting, pasting, moving around, and in general, manipulating whole s-expressions as individual units very convenient with appropriate editor support (e.g. Paredit). I know there are languages like Racket, Clojure, etc. which fulfil the requirement of extensive batteries included library + s-expressions, but they are much less popular than Python. In my ideal world, one of the top three popular languages would be a Lisp or Lisp-like language with an extensive standard library comparable to Python's.
3. Common Lisp for personal projects. Python for collaborative projects.
[1] https://www.lispworks.com/documentation/HyperSpec/Front/X_Sy...
1. Newspeak, Smalltalk 2. Smalltalk's code browser can be tedious, Newspeak's IDE is more coherent. TCPSockets, tools 3. Smalltalk's Seaside framework for server side web, Newspeak for clients.
1. PHP and C. 2. I wish C had PHP's native "associative array as junk drawer" data structure. Conversely, I wish PHP had more support for a "long lifetime" task that wasn't expected to vanish at the end of a page load. 3. Am I trying to target the web? Then PHP.
1. J, Lisp.
2. There are no such features. But I want all languages to have J in the same manner as they have regexp. I mean a tool like NumPy but with sweeter syntax and more possibilities.
3. J is for number munching, Lisp is for everything else.
1. Typescript, Ruby
2. Elegant typings, native to language syntax. Syntax flexibility.
3. Usually this choice is made for me or before me. And unfortunately it is often neither.
1. JS, Python
2. [i know u said 1, but there's so many] JS->Py: Looser type handling, nicer syntax for dicts. Py->JS: Less punctuation, comprehensions
3. I created a hybrid called Pith (https://news.ycombinator.com/item?id=46637033). JS with some nice Python features.
Appendix (Nov 2021)
A version of Python (call it "Pith" [0]) that fixes the annoying things, eg:
no colons
nicer dicts shorter dicts auto main function dict.get(prop) by default ternary elif? elf? what the helf ? W T F-string? len (like it's BASIC) a longer example i also like how JS allows you to call a function before it's defined (but Python doesn't -- why !!!)maybe add a do/while loop, multi-line comments ###, JS-style regex /reg/ instead of r"regex" blah blah blah
etc, etc, etc
it could be a preprocessor (like the C Preprocessor) that takes Pith and converts it to Python
for JS, take out parens, curly braces, semicolons, etc (make it "look" more like Pith)
i know these changes are fraught with peril, but i don't care (do i?). there's no reason (is there?) that i should be stuck with Guido's or Brendan's design choices [1] (i can make my own :) )
i even had an idea that you could write a program in Pith and it could output either Python or JS (or anything else). that might need a little more thought
[0] "Pith" is prolly already in use by something else, just humour me
[1] that's the bigger idea here -- take a language you like, fix all the things you don't like about it. maybe lisp with fewer parens
1. Python, C++.
2. Significant indents, Compiles.
3. Ease of coding, Performance.