#python-discussion
1 messages · Page 564 of 1
print(r"Hello\tWorld")
Hello
Try
print("|\v|")
hi bread
How are you today?
I will
And try
print("abcdefghij\r123")
just wrote an aimbot for osu
and against their ToS
are you planning to make it into a semi sentient existential crisis causing AI
no yes idk
Did you integrate llm
no-ish
Can he get banned for this?
yes
osu?
stay innocent kid
click circles to a beat game
is osu even multiplayer?
Osu ruins lives
other way around, I'm old AF
yes it is
thru leaderborads and rooms which use leaderboards
stay innocent old guy
<3
XD
so then why bring up a subject that we don't allow on the server?
let me flex a bit
it's like you want to be banned or something
who would want that?
maybe he just wrote it for practice
I am being very serious here that your conduct over the past many, many weeks has put you on thin ice. I can only warn you so many times before you just get removed from the server.
doesn't really matter as far as our server rules go. We can't judge intent.
you don't have to take me seriously though, i do know when to stop myself
Not really considering how often we've had to step in and warn you.
you wanted to do that though
anyway we could throw this under a rug
yoo do i do lists or if statements first
either is fine
I should write reverse pyinstaller
packaging an EXE with a Python x86 emulator and windows API interface
Just use docker
I should say that whenever anyone mentions pyinstaller
I already do
But you're welcome to join me
For a second I thought it was run Windows in a Docker container.
click, typer, cyclopts or fire fort a CLI tool?
There's an Australian fork of that. It's called Ocker.
lmao
i dont get it
I thought it was podmate
Australian? why not rekcod
bold move
im so confused on interpreted language vs compiled. Interpreted language is executed line by line whereas compiled is translated into machine code before running right?
It's a spectrum. Not either-or.
it is hard to categorise languages as being either interpreted or compiled
im confused
whys that
most languages you think of as being "interpreted" will have a compilation step
for example, Python compiles to Python bytecode
You're probably not at a stage where the distinction is very important
Python is somewhat in the middle ish of it. For example, while not being compiled to machine code, it's being compiled to a bytecode that the python VM runs
(although, this is an implementation detail of CPython, but that isn't really important)
maybe this is it
How much would you say you know about python at this point?
3/10? im learning cpp right now because thats what my courses uses
maybe 2, if i were to give it a rating
I'd say it's not worth wracking your brain over right now
Is worth me wracking my brain over?
probably. You're right to say that interpreted languages are executed by a separate program -- the interpreter -- whereas compiled languages are compiled to machine code and executed directly by the processor. I take issue with the "line by line" thing, but I suppose it is helpful to understand the uhh... philosophical difference
this is the most comprehensible answer ive gotten so far
thanks!
interpreted vs compiled is more of a spectrum. It depends on the context of the conversation and there also isn't a binary answer.
Some interpreted languages might still have a compilation step in the form of JIT
It also depends on the definitions of "compiled" people are working with
some people have a very narrow definition, others a broader one
Python is interpreted with a compile step and that step happens regardless of whether the JIT is enabled or not.
it gets complicated, and introductory courses will often provide a simplified view because it isn't worth introducing a load of confusing concepts just for a more technically correct definition
Does it have any implications for something you're trying to do?
i think you gave a perfect answer, thank you
The JIT portion potentially compiles though, no? That's the point I'm trying to make
Very rarely.
Right, but Python's compilation step isn't because of the JIT. The JIT is an optimization on that.
I did not say Python's compilation step was because of the JIT.
The bytecode compilation is not what I was talking about
SOMEONE please help me pick a cli tool for my CLI note taking app. Click, typer or cyclopts
I like cyclopts the most out of those options from what I've tried
"Some interpreted languages might still have a compilation step in the form of JIT" #python-discussion message
That's what I was responding to.
I like typer and have never used the others. I have no opinion on the other two
although I only tried typer for the briefest of minutes
Define the tool...
cli tool?
That does not say that the bytecode compilation step is because of the JIT. JIT is an additonal compilation step for hot code. You'll also notice a lack of Python mention in that sentence. Almost like it was on purpose
I don't know Cyclopts. I've found Typer to be easier to implement than Click whereas Click is more fully featured.
And I replied to that and said that that is incorrect for Python. That's all.
Single command? Multiple commands?
So, you're still misunderstanding what I'm saying but okay
The one time I saw code that used click, there were lots of stacked decorators, and I didn't like that
im going to be adding notes to json, with flags, im pretty noob sorry
I am not. You are misunderstanding me.
Can you guys please bicker in DMs?
Is this a single command with options or multiple commands?
Oh is that what DMs are for?
DM me to find out!
apparantly "Cyclopts is what you thought Typer was"
I'm just trying to clarify what I'm saying, since from my perspective Dan is responding to something I did not say. But I will drop the subject. I do think it's important to clarify misunderstandings for onlookers though.
I'd never heard of cyclopts, but I just read the readme and it looks pretty cool
We don't need to suppose that they're fighting. Both of them are level headed and they're just trying to reach mutual clarity
Preocts recommended me cyclopts. Very much prefer it over click
I think you're both talking past each other. But it's getting out of hand.
will be a single command like notes add "my apple was not very crunchy today, shitty JAZZ apple" --important --apples
I might try it out the next time I need a CLI interface. Usually I just hack something together with getopt
I have the feeling that they're all overkill and you should use argparse or a wrapper around it like dykes
Yes, clarification is important. But a long no no no no exchange just confuses onlookers. DM, reach consensus, come back. Maybe.
I saw it but the name scared me away
ugh, "CLI interface" 🤦
ill look into those TY dan
what about the word cyclopts scared you away?
Click, Typer, Cyclopts, etc. are great for multi-command applications (think git) while argparse is great for single command options (think cat)
It was a joke about cyclops
oh i will have many optional funcs like add delete modify etc etc, if thats what you mean
multi cmds are like git, which has git add, git commit, ... a bazillion others. If that' what you had in mind.
Then go with Typer over Click because, IMO, it has a smaller learning curve. (I can't speak for Cyclopts as I don't know it.)
yeh yeh there wont be just add, Thanks everyone!
It would be very easy to know it
Does anyone know the best places to learn python? I’ve been trying codecademy for a few days but was wondering if there were any better ways to learn python.
We have a list of resources here that we recommend. What don't you like about codecademy?
!learn
Here are the top free resources we recommend for people who are new to programming:
- Automate the Boring Stuff — an online book (also available to purchase as a physical book)
- Harvard’s CS50P course — video lectures (slides and notes provided) with exercises
- Python Programming MOOC 2026 course — text-based lessons with exercises
- Corey Schafer's YouTube playlist
For a full, curated list of educational resources we recommend, please see our resources page!
It kind of feels like flashcards that I’m using to study for a test 😅 I was wondering if there were more interactive and hands on ways to learn python
Python already is interactive. You should always be writing code along with whatever learning material you use
write, test, break, adjust
so making my own things based off of what I learn?
yes and even just playing around with the code to make sure you understand it properly
if you look at cs50p from the links above, it's a great course and it does include problem sets that you can work through
Alright, thank you
np!
why is list max 5
Codeacademy I have good things to say about, but sometimes the material is out of date or isn’t as in depth as I’d like.
it's not?
For Python I mean. Took their courses and were decent but I think still Python 2 assumed
car_brands = ['chevrolet', 'ford', 'chrysler' , 'dodge' , 'ram']
car_brands.append('buick')
print(car_brands[6])
We start indexing from 0
the last index of a list will always be len - 1
10 things in the list? last index is 9
472 things in a list? last index is 471
what you think
car_brands = ['chevrolet', 'ford', 'chrysler' , 'dodge' , 'ram']
car_brands.append('buick')
car_brands.append("cadillac")
print(f"My dad's first car was a {car_brands[6]}.")
holy progress
Great! As a neat trick in python, we can actually use negative numbers to index from the end of the list
so car_brands[-1]
the 4th line is not true my dads first car a Dacia
Nothing in this example
but sometimes -1 is really useful when you might not be sure how big the list will be
"I want the last element no matter the size of the list"
exactly
car_brands = ['chevrolet', 'ford', 'chrysler' , 'dodge' , 'ram']
car_brands.append('buick')
car_brands.append("dacia")
print(f"My dad's first car was a {car_brands[-1].capitalize()}.")
see how cool that is
I do
i put capitalize hella niche
quite commonly used*
i like coding man
What about reference the book? It explains very clearly!
this is some fun shit
it's a good feeling when you get it right
reference what book? Who is that message for?
for bleath
bleach?
yeah, you have just discuss the python commands?
What book are you referring to?
there are many, for example, "Introduction to python programming",
People come here to ask questions and discuss things though
If we all just stuck to our books and didn't talk to each other, this would be a very quiet place
It's already quiet enough with the rise of "the thing"
shhh not so loud
ok, i got it
oops 
anime = ['naruto', 'one piece (garbage)', 'dr stone', 'one punch man', 'bleach']
anime.append('jujutsu kaisen')
print(f"{anime[-1].title()} is my favorite anime, but i also like {anime[-2].capitalize()} and {anime[-4].title()} =).")
!run
where do you guys place imports from typing? (not in alphabetical order with the order standard lib packages, i assume)
nice program
first in each file
i ususally go typing imports
std library imports
3rd library imports
current project imports
I just do what isort does, which is alphabetical.
interesting. i was considering putting it on the bottom :p
but top or bottom both seem good
ah
I cause chaos by importing from typing before I use each type ||/s||
does isort still separate std lib, 3rd party, and project imports?
Yes, that's how PEP 8 does it.
gotcha
that sounds reasonable too then
How do I like print two listed items at once
What do you expect the output to look like for that?
does isort have any special case for typing imports?
Nah forget the program just how can I ?
I'm asking because I don't understand what you're asking. We would need to see an example of the output you're trying to achieve
In with the other imports, myself.
Not by default, though you can define custom sections if you want.
Okay look at my anime program what if I wanted to print anime [3 and 4]
like, alphabetical, or otherwise in the middle of std lib imports?
By themselves or as part of another string?
Themselves
Just alphabet with the stdlb imports. Eg:
from contextlib import contextmanager
import os
from os import fstat, pread, SEEK_SET, SEEK_CUR, SEEK_END
import mmap
from stat import S_ISREG
import sys
from threading import Lock, Thread
from typing import List
Third party imports separately below, then package internal imports below that if relevant.
it still sort of depends
!e
anime = ['naruto', 'one piece (garbage)', 'dr stone', 'one punch man', 'bleach']
print(anime[3], anime[4])
:white_check_mark: Your 3.14 eval job has completed with return code 0.
one punch man bleach
you can just use a comma to print multiple things
if you want another list of those 2 anime, we can use slicing
!e
anime = ['naruto', 'one piece (garbage)', 'dr stone', 'one punch man', 'bleach']
print(anime[3:5])
:white_check_mark: Your 3.14 eval job has completed with return code 0.
['one punch man', 'bleach']
the standard form is import from stdlib (alphabetized), then from import from stdlib alphabetized, then third-party, then first-party.
got it got it 👍
cameron's style is very unusual.
It's entirely a personal preference.
from typing after import?
i put from typing first
You need to be odd to be number 1 
Aightttt
ah this is great to know regarding from imports, i wasn't clear on the standard for those before
I’m defending Homer’s relative
use ruff, it will arrange your imports
i have been separating them anyway
i dont like linters formatting my code
yes 👍 i think i'm going to keep my typing imports below all the others
because of personal preference
if i use an autoformatter that puts it in with the other std lib imports, though, i won't stop it
sure, don't try to stop an autoformatter
actually i did just get ruff
i'm having fun trying new things like ruff and ty (i've never used type annotations before)
the philosophy of black seemed respectable so that's why i thought i'd like ruff for formatting
thats reasonable
i do it the same way, more or less, but from first, then import
thanks, looks pretty clean
Imports are always put at the top of the file, just after any module comments and docstrings, and before module globals and constants.
Imports should be grouped in the following order:
Standard library imports.
Related third party imports.
Local application/library specific imports.
You should put a blank line between each group of imports.
shameless utm_source
i guess typing technically is included in standard library imports
wasn't that text written before typing existed though?
Why would typing be spcial?
yea i asked chatgpt if there's a style guide for this
i looked at django, pandas, flask
flask is import then from
isort seems to do import then from
in the pep8 page i didn't find anything in that regard, just that paragraph
I put my typing imports in a TYPE_CHECKING block
the pep8 states the order, but not if import is before from
i guess a couple things different between it and other packages are that it doesn't affect how the program functions, and that you might generally import it in every module in a project that has type annotations
hi
from typing import TYPE_CHECKING
if TYPE_CHECKING:
from typing import Literal
have you guys used pygame
it seems reasonable to separate logic/imports related to typing and logic/imports related to program functionality
I'm sure some here have used it. Did you have a question about it?
i dont think thats smart
hello
could you try out my game?
you wouldn't be able to do instance checks or anything like that (with those types)
You could try posting it in the #game-development channel if you're looking for people to try it out and give feedback
Right. If I were needing to do instance checks then I would import typing outside of that block
why importing it in that block to begin with
Copying what I've seen other people do
🤔
is this not going to give you runtime errors without deferred annotations
That's why I defer annotations
you normally use that when you want to import stuff that normally, can't be imported at runtime, but only used for type checkers
Right. And I don't use the typing module during runtime typically. If I were to, then I would move it outside of the type checking block. In fact that's what I do in one of my files where the typed are required at runtime
TYPE_CHECKING's value is False at runtime, and True to type checkers
I know that
uh alright then
i wonder if you'd analyze the typehint at runtime what would happen
I don't do isinstance checks against stuff in the typing module typically. If I needed to, then I would move the typing import outside of the type checking block so I could use them at run time. I also defer annotations. I do this so I can use new typing features in older Python versions without needing to use typing_extensions. This works, and I've tested it against each Python version in ci
from __future__ import annotations
if TYPE_CHECKING:
from collections.abc import Callable
def a(b: Callable):
...
from typing import get_type_hints
print(get_type_hints(a))
it errors
name 'Callable' is not defined
thats smart, i see
i think that this is impossible at runtime
you can get the name in annotations, but not the real type
vegetables = ['celery', 'tomato', 'broccoli', 'onion']
del vegetables[1]
print(vegetables)
I forgot tomatoes are "fruits"
i wonder whats the reasoning behind TYPE_CHECKING to begin with
SupportsKeysAndGetItem for example
thats funny
in what sense?
why not make them importable
whats the need for this _typeshed
For importing types that aren't real
something something Conway's law
Typeshed types don't exist at runtime, hence needing to import with typechecking
i understand that, the question is why
I really wish they implemented TYPE_CHECKING being in builtins without needing to import
you can make them real types
Py3.15 lazy imports will probably remove the need for it
People maintaining typeshed are a different set from people maintaining Python, and typeshed moves at a somewhat independent pace. So it could just be that people don't want to bother arguing about including these things in Python at runtime
Copy them into your own codebase
You're correct
Yeah, I think it's a bit silly to have imports from typeshed. Just vendor them into your program/stub
so .pop( ) is the same as [-1]?
No
pop removes the item from the list
the last item ?
the solution is that you do this typechecking thingy... but that solution has drawbacks
not a very good design in my opinion
yea but not everyone is gonna do that
yes, but we don't use negative with pop
if TYPE_CHECKING:
from _typeshed import ...
is a common pattern
so get_type_hints can't be 100% reliable
since it can error... a situation i didn't think of
the whole idea of if TYPE_CHECKING is weak design choice in my opinion
perhaps i should try to stop using it at all
from typing import TYPE_CHECKING
if TYPE_CHECKING:
from _typeshed import ...
What situation is that?
^
the type doesn't exist in runtime, therefore it can't be resolved
it's an interesting approach
Note
If Format.VALUE is used and any forward references in the annotations of obj are not resolvable, a NameError exception is raised. For example, this can happen with names imported under if TYPE_CHECKING. More generally, any kind of exception can be raised if an annotation contains invalid Python code.
Yeah, I'm just not going to argue about that whole thing
It's not worth it
IMO the design is fine.
Assuming you can construct proper values from type structures in real time is not as simple as it sounds
if the type exists at runtime you can grab it
Linear execution is fundamentally exclusive with cyclic type definitions. You can't have both
Unless you want to run a resolver of some sort (and/or re-execute module imports with TYPE_CHECKING on), there's not much you can do here
not sure what u're saying
when TYPE_CHECKING is on, that block doesn't get executed in runtime
The purpose of TYPE_CHECKING is twofold:
- to break cyclic dependencies introduced by types referencing types defined after them
- to avoid importing heavyweight typing instances when not required
class SupportsKeysAndGetItem(Protocol[_KT, _VT_co]):
def keys(self) -> Iterable[_KT]: ...
def __getitem__(self, key: _KT, /) -> _VT_co: ...
whats cyclic about it
then why its in _typeshed
!pep 781
there's trucks and cars. You have to pick one. TYPE_CHECKING allows you to selectively decide "I want a car"
maybe we'll see it in 3.16
The rest is down to the implementer
We have lazy imports at home
i think that your reasoning might make sense in some scenarios, although i'm not sure which
yet it can be problematic
not to change the topic, but how do you all document module variables?
what you mean by that
Module variables as in globals or module variables as in constants
Triple quote on next line
someone hire me bro
like how to write docstrings for a variable (x = 5) defined in a module that can be imported
!rule 9
x = 5
""" :) """
uh i see...
Does this work?
👍
i've seen this and it looks great to me
i've also seen
#: documentation here
x = 5
do you by chance know if any of these is more standard?
Its not standard, but Pyright supports it
oops
And mkdocs
Comments are more standard, I believe, if only because I've seen them more
thats interesting
well, i dont do any of this, so i dont know
might be a dumb question, but, when do type checkers look at docstrings?
this is good news for me 👍
There's also annotaed docs. ```py
x: Annotated[Literal, Doc("")]
^
Is it a constant? IS it a class?
uh, those i've seen
now i'm curious if pycharm recognizes any docstrings for module variables actually, i'll check
Maybe I just need more sleep, but I'm not sure what you mean by module variable in this context
The lsp needs to know what function you're calling to show the documentation.
a constant would be a fine example, yeah
note the pyright doesn't support this
from what it seems, only unalivejoy example is supported
i didn't realize pyright handled displaying of documentation too 😅 thanks. i thought it was just about adding squiggly lines when types didn't match
I think Ty supports it too. Not sure about jedi or pylsp or rope
and i didn't know about it, its new to me
Hello
I don't think there's a universal suggestion, but I've seen some strategies to just use a regular """ """ afterwards, do a # comment before, or make it part of an enum so docstring shows up but imo that can be a bit misleading
hi there
pycharm does support triple quotes as docstrings for module variables 🎉
Yay!
Just make sure that whatever doctool you're using (if you're using an autodoc tool) is compliant with it
Annotated is also a fine solution
Not enough for the pep to be accepted
yes! i'm planning on using mkdocs so i think i'm in the clear
Don't.
oh
what pep
!pep 727
is it like a more modern version of mkdocs?
Yes
And the new mkdocs maintainer has plans for 2.0 to remove plugins
(Lovelydinosaur)
uh
i think i need a latex plugin in my case
is c a good first lang
Python is a good first lang. What do you want to build?
so there may be a high chance i will try zensical
I think the learning curve is too steep.
no idea
Why are you learning python?
why not
its better than doing nun
iirc, either you or godly suggested that c as a first language makes it likely a last language
I suppose that's a good reason
There are benefits to learning C as a first language, but it is pretty rough.
Why this over Sphinx?
no reason that i can think of
i dont have time to learn py after school i open codium and its already 10 pm
You can use sphinx if you want.
uh its a phrase i mean...
ima need to lock in this weekend
godly might have worded it that way. I've said that I think people who start with C are more likely to give up programming entirely.
i know i'm just agreeing
you're not invested in mkdocs yet right haha
ooo
nope, haven't tried sphinx, mkdocs, or zensical yet
The benefit to learning C first is that in academia going for a CS degree often has you learning the fundamentals and building up, and C is a language that does not abstract all those fundamentals away, so it's useful in that regard.
That being said, there is a lag/delay time between you learning to code and actually being able to create something meaningful.
I think, to gain and keep interest, first language should be something else but C is a good language to use when you're actively learning about CS fundamentals.
i wasn't quoting but paraphrasing, and in this case, i was apparently remembering your comments
-# unless you both believe the same thing
you can even try them all to be fair
But don't tell them that you're using other autotooling, they might get jealous and it leads to drama
from reading online, somewhere i got the impression that sphinx is "getting old", but i can't comment on it at all myself
What are you trying to accomplish? Is this for a project or for school?
Sphinx is... fine.
self growth
learning
@chilly whale I'm exercising directorial privilege to ping you and ask if you think C is a good first language.
Cool. Python is certainly open for that.
Someone's going mad with power!
omg did I ping him too >.>
you didn't. replies only ping the person you're replying to.
!learn
Any language will do, @tame raptor . Python has a lot of resources available if you don't already know what you want to do. Other languages don't have as much support there.
Here are the top free resources we recommend for people who are new to programming:
- Automate the Boring Stuff — an online book (also available to purchase as a physical book)
- Harvard’s CS50P course — video lectures (slides and notes provided) with exercises
- Python Programming MOOC 2026 course — text-based lessons with exercises
- Corey Schafer's YouTube playlist
For a full, curated list of educational resources we recommend, please see our resources page!
Is that a new thing?
no
Because they certainly used to
must resist extremely relevant simpsons gif
true!
it was new several years ago.
I still remember the Lemon fiasco
yeah, that was fucking hillarious
but to be fair i know sphinx is considered very good
Objectively hilarious.
So hilarious. It's still pinned
maybe I should reenact it but with 420k people instead of 90
C's not the worst first language, but it's not great. It does a terrible job of helping you figure out why your program isn't working.
There are definitely worse languages than C to learn first
oh, absolutely. But there are definitely better ones
what was that
i claimed in here btw, before, that i didn't understand C# properly until i learnt C
Only admins had everyone ping permissions, so a new user tried to ping everyone, but lemon responding telling them to knock it off DID ping everyone
#ot1-perplexing-regexing message
4th-to-last pin in ot0
@robust ledge fyi while i was in class i was like i cant wait to go back home to hop on codium
#python-discussion message
lemon replied to quoted someone who tried (and failed because they lacked perms) to ping everyone. but because lemon was a director, it did in fact ping everyone.
You can still see the massive spike in people leaving the server
I'm not saying not to learn C. I think it's a very good language to know. I don't think it's a good language to learn first
when looking at historical data
👍 it's the one i've always heard about, except until recently
Learning C is like.... you already know you like programming and want to stick with it
So I guess I'm in the "not first language" camp
🤣 🤣 🤣 🤣
If you don't know if you'll like mario kart, don't play against the person that tries way too hard to humiliate you for your first game
and C will humiliate you
🙃
That said, if you already know C, you can ramp pretty damn quickly to other programming languages. The C -> Python ramp was remarkably smooth
although switch/match case still trips me up
what is pandas
The best animal
and best data science library
and a 3d gaming library that I've never used
A data science library for handing, well, data.
Does zensical work with rst files?
pandas live their lives with wild abandon
no idea, never used it. Sphinx has been enough for projects where I've needed to care
sp[hinx for most modules, then for the important stuff I just did it myself
directly
one of the funniest things i've seen in a while
don't forget that pandas are bears. and how you would react if you saw any other non-panda bear.
How many members left the server the day Lemon pinged everyone?
I'd have to ask joe
Pandas are my spirit animal. Or fursona. Whichever is more politically correct
Their experience just resonates with me
just randomly falling from trees, eating shit as they go down slides, etc
i think that Python is a better choice, but C# for example, is a worse choice
i think OOP confused me, C got me to be strong on the core, understand how things work, and even i appreciated more the normal for loops, while loops, the very basics
is python docs a good way to learn py
I would not recommend C# as a first language
my city did a public transit rewind with stats about your use of the system (based on the serial number of your transit card), and it assigned everyone a mascot based on their usage profile. my transit fursona was an explorer dog.
!learn
Try some of these resources to start.
Here are the top free resources we recommend for people who are new to programming:
- Automate the Boring Stuff — an online book (also available to purchase as a physical book)
- Harvard’s CS50P course — video lectures (slides and notes provided) with exercises
- Python Programming MOOC 2026 course — text-based lessons with exercises
- Corey Schafer's YouTube playlist
For a full, curated list of educational resources we recommend, please see our resources page!
I do think python is a really good first language. Because it shortens the friction between trying out programming and actually being able to see something that you've made
But also.... it becomes obvious if you've only used python to program when you're discussing CS fundamentals
and it's quite good at helping you understand why something didn't work
That is true, the error messages tend to be more helpful
don't underestimate the value of seeing a traceback automatically when your program has a bug
Yeah
It also has an absolute massive community to support learning. Project idea, tutorial, and the like.
"What do I program" is a question that can be tough for those interested but without a clear goal.
the worst part, i think, about using c or c-like languages is the absolute garbage support for basic data types.
"uh hi, how do i make a string? also i want to know how long it is"
"turn back now, dear newbie. there is no path forward but to suffer. such is the way of the c strings"
idk how to say this but i dont like these
What don't you like about them?
i wanna learn with the documentationnnn
I think I just.... my "problems" with python are similar with my problems with vibe coding, it helps people make things where they have no idea what's happening or how it works. Having to work with people like that can be frustrating when they're also stubborn.
What will you build with the documentation?
idk waht to build
That's what the resources I suggested will solve. Until you have something to build, the documentation isn't much help.
aye, sometimes i strongly wonder if maybe python is too easy
I'm not trying to be gatekeepy. But it is genuinely frustrating working with someone who looks semiqualified on paper but has no idea what they're doing and refuses to learn.
checkmate i guess ill use many sources of the one you gave like automate the boring stuff and the harvard thing
so 2 sources of learning
That's a good tactic. Many resources can give you more information than one.
!learn
Here are the top free resources we recommend for people who are new to programming:
- Automate the Boring Stuff — an online book (also available to purchase as a physical book)
- Harvard’s CS50P course — video lectures (slides and notes provided) with exercises
- Python Programming MOOC 2026 course — text-based lessons with exercises
- Corey Schafer's YouTube playlist
For a full, curated list of educational resources we recommend, please see our resources page!
i agree, to be fair, maybe i'm not even sure if python as a first langauge is bbetter than C
Yo Preocts, wanna feel like you're back on helpdesk again?
i guess i dont really have an answer for this one, since Python wasn't my first language
You could even just use Automate the Boring Stuff for "what to build" and the documentation for "how to build". Each chapter has small projects to complete.
This isn't a helpdesk?
Maybe it's more of a problem how these languages are taught? Idk
programming didn't really sit in my mind until i learnt C, my second language i guess
#ot2-never-nester’s-nightmare is playing https://servicedesk-simulator.com/
It's giving me a panic attack
what do you think of python crash course the book
Did someone publish something about that book?
Honestly, early in my life I sort of just tried like..... every language. Then engineering academia forced Matlab, then cs courses forced C, then a job I got forced python, lol.
Personally I think python then rust is a really good order, because as discussed python is a very good intro. And rust introduce you to the concept of lower level memory management, but still with actual good abstractions
I've seen like it five times this week
I might pick up rust. But also, C is still incredibly valuable to know
i got yer simulator, right from that website's front page
"of course i restarted the computer. twice now in the past 15 minutes"
computer: UPTIME: 91d 21h 45m 55s
What about Zig?
the simplicity of C makes it a very good language to learn
i wonder what happened to A and B \s
To clarify, I'm speaking most specifically to the job market
yoo
you should
asm is simpler than c ya know
There's one position in my industry that is looking for an algo dev who knows both Python and C++ (in the life sciences) and I've gotten literally 20 recruiters reaching out to me about the same position
It's ridiculous
simplicity?
I think it's a book that many have used. So long as it gets you writing code, it's a good book.
Easy fix. Don't play it.
yeah i've seen a lot more rust in industry than i have zig
i'm talking more about... getting programming to sit in your mind
Yeah, like... how good a programming language is I consider somewhat separate than how hireable it is
it's a really small language with not a lot of features (which is why it's simple). doesn't necessarily mean it's easy, though.
i found C to be very useful for that
i really hate asm
Maybe one day I'll think of languages outside of the direct context of how much it contributes to finding a job, but today is not that day.
apparently unis used to teach their intro to programming courses in lisp
it'd be cool to bring that back
a doesn't exist. c came from b, which came from bcpl, where the b is short for "basic". b is nearly dead outside hella legacy stuff
I think either b or vb was my first language
feels like i know nothing then lol
like decades ago
b stands for basic?
Computer Science courses experiment with various "first languages".
Yeah
I learned it at a summer camp
That summer camp was so fun because they had bought starcraft and you could play starcraft in daycare after the lessons were done
The b in bcpl stood for basic.
My first language was BASIC, then Z80 assembler.
Then at uni Pascal and C and Bourne shell. Then other things....
Which one
MIT i believe? im sure there were more
C isn't an acronym though
Probably Racket then
I coded pid controllers in ASM. It was a fun challenge, but I don't think a particularly like.... usable skill
oh, you were talking about what kind of lisp
like it was educational and is a cool flex
whats a good vs code theme
The class that had you doing systems stuff in ASM was a masochist class most people failed, lol
wat
are you actually serious
seems like it is some sort of scheme
pid controllers with essentially 0 latency was pretty intense from a physics perspective
speaking of racket, i think https://htdp.org/2026-2-25//Book/index.html is a rather pleasant "intro to programming" book.
asm might be too low level, but coding PID controllers is hardly unusable
Oh, agreed
too low level is the definition i guess
that sounds fun :-)
also, asm is very different than any other language i know
i like gruvbox material
My point was that ASM is not the optimal language for it. You can code faster in C and the memory/speed is not meaningful enough to suffer through the pain
no common syntax, no common attitudes
I don't have the required physics or math to know how to program a PID
Honestly, it's sort of misleading to describe ASM as a single language at all
wouldn't a c compiler spit out faster machine code than assembly you could write in the majority of cases anyways
asm is never fun, stop lying
yeah i'm with mtn here. i've done asm. meh
I thought you were commenting on relationship status at first
the worst type of typo 😔
Doing PID in ASM was mostly useful as an exercise in understanding how they worked, how registers worked, etc
SMB and other NES games were programmed entirely in 6502 ASM
yeah like ultimate said, "assembly" just kind of describes all of the languages that are more or less 1:1 to their machine code instructions for a particular architecture
they almost found about us
heyyo ult: as asm fella, thuri has asm question:
MOV SRC DST or MOV DST SRC?
Are you trying to bring back the nightmares
i'm trying to judge you
@dusty ember you know ball this theme is nice
so it makes sense that there's not a common syntax, because you aren't working with common instruction sets.
I haven't touched ASM since academia
I didn't see much sunlight when I was taking that course. So much time in the lab from early morning to when the sun was already down
yea, but i mean, comparing asm to C, or js, or java, or c#, etc, its not similar to them at all
ASM is lower level than pretty much anyone will ever need in the modern day
well, yeah, you're comparing what is essentially machine code to "real" programming languages.
It's lower than Assembly?
yea, i guess
alas, ult of chaos shows not knowlege but wisdom, and avoids answering the trap question 😔
ASM is short for assembly
Oh, I think people should code in it at least once in their lives
Hah, we had a group project in 2nd semester (4th for me 🐒 ) where my group had to implement Huffman compression in handwritten assembly (though we were allowed to use C for I/O, as well as for the option parsing via getopt).
That was... interesting
It could be an eye-opening experience
It has been a while, but is it just asking the difference between x86 vs ARM? Well, not specifically that, but assembler dependent
I really hated group coding projects in college because I wound up doing them all myself
I mean it was fine ig, I got the grade
what level of type checking do you guys run?
i usually run pyright strict
alright, cya guys, thank you for the conversation, was fun
I worked with ARM in academia, but found that most jobs these days that mention microcontrollers are talking about the intel/x86 architecture
!!!
my project worked fine, but as soon as i turned on type checkin gholy it didnt like what i was doing lol
MOV SRC DST is part of the otherwise-awful sigil-addicted at&t style, MOV DST SRC is the abomination in the otherwise-okay intel style
both for x86
The answer is that I flipped over the assembler manual and looked up the syntax it needed 😂
That particular one I just mentioned was actually pretty okay. We were a group of 3, one of the guys was a friend from school who started with me, the other guy was a random, but tbh the random was probably the best of our group.
Unfortunately my friend apparently had a pretty shitty implementation. About 1-2 years after the assignment I took the code, and re-wrote his part in pure, relatively naive C, and it was about 50x faster overall :/
All of my code passes mypy strict and ty
isn't dst := src the way it works in high level programming languages?
<insert admiral ackbar trap gif>
Did you have type annotations in the project? Were you running with less strict type checker settings prior?
that doesn't have the mov part, though. imagine swapping the order for something like os.rename
Now my question to you is how you feel about command line copy source destination
its was OFF before but the project was running fine. Its very small and basic. But i turned it onto basic and had to fix some stuff like add the str and int to body and tags:
@classmethod
def dict_to_note(cls, d: dict):
return cls(
id=d["id"],
title=d["title"],
body= str(d.get("body")),
tags= list(d.get("tags")),
priority= str(d.get("priority")),
creation= str(d.get("creation")),
done=d.get("done", False),
pinned=d.get("pinned", False),
archived=d.get("archived", False),
)
though tags is still giving me Argument of type "Unknown | None" cannot be assigned to parameter "iterable" of type "Iterable[_T@list]"
You need to decide if you want type checking or not. It has both benefits and downsides.
if this is an editor question, i don't use either vim nor emacs
ohhh, so its just for like, cleaner code? im a bit confused
I admit that I still have to lookup the order things come in for a lot of module functions that require two arguments. 3 or more is fine, it's when it's just two that I'll sometimes get it mixed up
Is it sane to use typing as like.... annotation more than enforcement?
I mostly just want docstrings to be clear and obvious
and signatures
The three main reasons people use type annotations:
- prevent certain classes of bugs
- precisely document interfaces of functions/methods/classes
- allow for better refactoring and navigation
For example, in this method you're passingd.get("tags")intolist().dis adict[Any, Any], sod.get(<str>)is of typeAny | None. The value can beNone, butlist()only expects an iterable, not None, so this is a type checking error. (and it happens to show a real problem that will happen at runtime)
The main downsides:
- you need to learn the type system -- the model of python that type checkers understand. It gets quite complex and the resources for learning it from scratch aren't great
- you might need to adjust a lot of bug-free code because your type checker doesn't "understand it" (since it has a limited number of supporte dpatterns)
- you might not like to clutter the code with annotations if you don't like them aesthetically
Yeah, I'm in category 2. I don't really use type annotations when I'd be using lots of Nones or Anys or whatever that just isn't clear
thanks fix error thatys great info. For now, and this project ill keepo it off. I guess it can be project dependant
4: make the ide shut up
It is a bit controversial, but I would recommend against using annotations without type checking. You will inevitably annotate some functions incorrectly, and readers will be confused (if someone else is going to read the code).
But then they have to keep you employed so that you can answer questions about the code, guaranteeing your job security ||/s||
They told me to write secure code. They didn't specify what kind of security.
Isn't it more like it's gonna speak more
Cause it's gonna warn you
not having sufficient and correct type hints can induce a ton of warnings/errors if type checking features are turned on
you can always turn them off, of course
i dont think anybody should be making new python code without type hints
it's ok to write code without type hints. the real problem is writing type hints but not checking them.
why write code without type hints? i assume the developer knows the type when theyre developing, why not annotate it?
perhaps it's clear enough without them. We wrote Python for 20 years without type hints. Perhaps it's not a program that will be kept for very long.
sometimes the types are not simple to express
i have a real life example i think, which is functions that represent mathematical functions
you might want to allow them to be used with normal python numbers, but also numpy arrays, and also other kinds of arrays (tensorflow, etc.), or maybe arbitrary precision numbers, or symbolic objects like in sympy
i think its rare that thats the case for the majority of types in a program, i understand some cases can have complicated types but thats not a reason to not type the whole program - using mostly typing + some any is better than no typing
(in my case, protocols might help, but they weren't perfect either)
slows me down
i agree that type hints improve most programs
my IDE is going to call me out on type clash issues anyways
negligibly
I'm in the same camp as ned. Of course they will improve the program, but it's not like code without it is unusable/unreadable
in general, i'm still a bit confused on how to reconcile duck typing and type hinting in python
def foo(param: duck)
i try this but the duck always flies away
i need help with a project a friend made for me but gave me the source code to, i don't know to use it.
you reconcile them by using type annotations that describe how the thing will behave, not what it exactly is. A function takes a list? Probably use Iterable as a type hint.
Didn't they give you instructions?
i think overhead of adding typing info (to new code) is worse any future pain from no typing, to me adding type hints is just a part of creating a function signature
no, he just gave me his "source code" to a tool, i've never used python too know what that means.
it's great that you find them natural and useful.
Why would your friend give you code that you don't know how to run without instructions?
perhaps the instinct stems from other languages you know
What does the code do?
Hello
for typing python? i think the instinct is just from writing typed python by default for a few years now (hence why i would encourage others to do the same)
i struggle to write javascript nowadays
see that's where you went wrong
same for python without type hints
you have to write typescript
why struggle to write javascript when you don't have to write it at all!
typing has become such an essential part of my programming habit
i had a period where i kept trying to add typehints to js code, but realistically writing js is always a pain because of how awful the language is
for an interview though
just let me use typescript
just write typescript then delete all the typescript parts after
i actually started writing jsdocs in js code and hoping my ide would give atleast some typing info when i was calling things
yeah I have done that before
typescripts typing system sounds really cool, but i dont think ill ever get to use it
part of the appeal is that your lsp helps to catch silly mistakes
using the typing info
though i guess that probably speaks as to how messy i am as a programmer
everyone makes silly mistakes
i genuinely think that it would be easier for beginners if you help them set up their lsp properly first
say the function is
def add(x, y):
return x + y
i think i'm confused on where to draw the line between:
- the most general type hint, which i think is using
x: Addable, etc., whereAddableis anything that supports addition, and which i believe will flag the same arguments that would cause errors at runtime - and something more restrictive, which would indicate what the function author intends
xto be (e.g. something number-like but not a string)
if anyone could help me use this source code thatd be awesome
What is the code?
you may want to post it in #1035199133436354600 if it's long
its a "remote access trojan" but theres no malicious intent it's purely a project.
yeah we're not going to help you with malicious code even if your intents are good
!rule 5
5. Do not provide or request help on projects that may violate terms of service, or that may be deemed inappropriate, malicious, or illegal.
alright well, then could you tell me how to make a source code into a executable.
thats not against any tos.
@harsh field It seems you have been met with a refusal. Sorry, it was right there. 😁
That's not really something someone with zero python experience will be able to do. We're happy to point you to some beginners resources though if you'd like to learn more about python
take your life
well
I may live long or I may not. Patience is a virtue.
it's the python discord server?
!ban 1501644635565592676 asking about malicious code and unacceptable messages to another member
:incoming_envelope: :ok_hand: applied ban to @harsh field permanently.
like speeding in front of a cop
how do i make a nuclear bomb?
we dont help with that
ok can you teach me how to create tritium, that's not against the rules
back to python
writing malware in python is an interesting take
"my friend gave me this code" is usually slang for "I downloaded sketchy code online"
writing malware in python is an interesting take
i've also heard the idea that type hints for public functions should be more general, since it's harder to anticipate how the functions will be used, while those for internal functions should be more precise
like how open accepts bytes, str, or pathlib objects
nothing makes me happier than when I'm working with a new library and it naturally accepts pathlib objects
i think so, yes
i mean... internal functions can also be more general, if the function works for children objects, right?
i'm not completely sure what you mean by children object
like objects of classes that extends some parent class where the function is meant to operate on the parent class
idk if im making sense
i havent written python in a few weeks
Hii
as the author of the function, you can decide what types it will accept.
but like the author could have this idea as a guideline to help make them decide how to type the function
in my case, i'm thinking of a mathematical function again, like add(a, b). if it's only used internally and always with floats, the type hints for a and b could be float. but if it's part of the public API, you might not want to restrict others to only using it with floats
what's going on everyonee
true, you might not want to limit it, but that's your decision as the designer of the API.
ah i think i understand. i think what i mean is that if the function is just used internally, it will be clear what classes will be passed to it. there's no need to anticipate the "public" making their own new class that you haven't included in the type hints
the built in + already dealt with this problem for you
this makes sense. part of my issue might just be that i'm dealing with some functions arguments that i might want to be very general. i suppose in this case i can just use a very general type hint, then. and in other cases, they can be more precise
this reminds me of an older post on the python subreddit that plays devil's advocate against type hinting, specifically showing issues with accurately yet fully hinting
def slow_add(a, b):
time.sleep(0.1)
return a + b
``` noting, to their pain, that more than just numbers can be added together
true
"anything that implements __add__" typehint
yes! but in my irl case, add is some more complicated math function that isn't built in or in another library
omg it's a trait
rust reference!!1!1!!1!1!
that's attempt 4. they stop at nine
I think I just won't typehint and save everyone the trouble
note that in Python's type annotations system int is treated as a subtype of float
surprisingly
the top comment is that - essentially "if it sucks this hard, then just use any"
ive noticed that if my object is hard to type, it's probably because the data structure is wrong
i need to sit down and read this thoroughly, the first part sounds a lot like what i'm going through
(this is my current decision 😮 maybe it'll be the last)
ah, pandas. how useful ye are, yet how horrifying thine types be 😔
one of the downsides of lambdas, too 😔 can't put in type asserts
proper closure when
try-except expressions when
everything-is-an-expression when
no one has ever proposed syntax for lambdas that allow more than 1 expression that ever got much community buy-in
pythons lambda syntax is horrible anyways
yes 👍 this has worked out nice for me at least, it seems like a reasonable choice
sometimes i wonder if adding lambdas would be rejected if they weren't already in
i think the current syntax would be rejected, at least
theres probably a few justifiable cases, but the current syntax makes them sufficiently obnoxious that they arent used much else anyway
i hope
python 4: removed lambdas
hm? They're used quite a lot
key= my beloved
i personally dont use lambdas
how does it feel having never called sorted or .sort in your entire life
a quick, unscientific grep for lambdas in the stdlib finds just shy of 2500 of them
$ git grep 'lambda.*:' **/*.py | wc -l
2436
sorted and callbacks mostly, however i try to not use them at all
Results for any search with code search are restricted to 100 results (5 pages)
bruh
why does github add a magnificent feature before severely restricting its use and then crippling it
istg if the slop floodbots are ddosing the code search 😡
yea, but restricted to a few specific classes of use cases i assume?
I don't really know what you mean by that...
they're used whenever you have a function that needs to be referenced at only one place and whose body is a single expression
as in sort keys, filter keys, and some callbacks, whereas in other languages its typical to use a chain of anonymous functions when working with things like lists using .map
Python would use list comprehensions or generator expressions in that case instead, typically
yea ig im just trying to express my dissatisfaction with how hard it is to express some functional concepts in python
list comps are functional
It's weird how "functional programming" somehow encompasses two pretty unrelated ideas, immutable data and passing functions around as values a lot
i wasnt a fan of functional programming until i had to parse csv data and turn it into useful results, the interface for just doing things to lists is amazing
If you're new to the language, we have some beginner resources here you can check out
!learn
Here are the top free resources we recommend for people who are new to programming:
- Automate the Boring Stuff — an online book (also available to purchase as a physical book)
- Harvard’s CS50P course — video lectures (slides and notes provided) with exercises
- Python Programming MOOC 2026 course — text-based lessons with exercises
- Corey Schafer's YouTube playlist
For a full, curated list of educational resources we recommend, please see our resources page!
In Python, what's the single most trivial way of quickly numerating all directories which $USER has permissions to?
Stressing on quickly
I'm thinking subproc + find, but is there anything cheaper and just as quick?
Well, "quick".
You check each files and folders for permissions lol
That... sounds like a bad idea, and rather expensive
Which I'm trying to avoid
Or you could ask the OS for the info about the files especially the permissions
You can easily get those information
...Okay? Directories have permissions and one could find out, but you're still not answering my question
I'm crawling through a very large HDD + NFS.
So, I'd like to minimize syscalls + stat calls
Naive approach has been running for the past thirty minutes and I don't have the time to wait that out
good morning!
'Figured it out'
what did you settle on?
import os
from concurrent.futures import ThreadPoolExecutor, as_completed
def scan_dir(path):
subdirs = []
try:
with os.scandir(path) as it:
for entry in it:
if entry.is_dir(follow_symlinks=False):
subdirs.append(entry.path)
except (PermissionError, OSError):
return []
return subdirs
def crawl(root, workers=8):
results = []
queue = [root]
with ThreadPoolExecutor(max_workers=workers) as ex:
while queue:
futures = [ex.submit(scan_dir, p) for p in queue]
queue = []
for f in as_completed(futures):
subdirs = f.result()
results.extend(subdirs)
queue.extend(subdirs)
return results
eivl....the project ,that is given to me, is very overwhelming
It's pretty much that, though Google says os.scandir is much cheaper than iterdir
if you count what? sys calls?
build something else instead
And regular Python overhead, yes
os.scandir apparently returns you a DirEntry straight from the OS dir stream, and said objects cache metadata from readdir
entry.is_dir skips an entire stat call
do you know any good place where i can find projects?
what happends if you try out ProcessPoolExecutor or InterpreterPoolExecutor instead?
That would be a great question, I don't know
What are your other interests?
the ones that are good are the one you make yourself
depending on how you read and use it, there might be a slight difference. sometimes you expect that the overhead of a process pool is to large, but it is better... but for IO operations, threadpool will wiin on paper
so make something related to that. Some sort of equation solver or maybe even a visualizer
Think of something that you might personally find useful
Yeah, it's strictly I/O
okay ..i will try
I suppose it would win out if I was doing anything other than enumerating them
when i did the 1brc i tested with multiprocessing pools and threadpools and found expected results during the whole process
Say, hashing or parsing
the 1brc is all about parsing text
Several of us
i am actually out of ideas
i want to make some tool for my uni people cuz its very easy to advertise it
like there are already alotof them like attendence tracker cgpa tracker
i was working on a ranking based website of leetcode but someone else did it before me
now i am looking for ideas
does ur uni have some amazing tool? plz do tell
Probably, ask you question to get responses:)
hiii
hello
someone else did it before me
So?
bro does your uni have some amazing tool?
just dont mention dating platform or something 😭
I'm not in uni...
i see
how about an AI-use tracker
tracks how much time or how often you spend asking chatgpt or ai
actually a great and unique idea
thanks sir
btw incredible user name
i just had my applied physics exam yesterday 😭
thanks and nice
was it the final?
Hey guys!!
Can you tell me what to do first DSA or OOP
welcome
Whats ur aim?
as far as i know they're both great to know
hello
i did dsa first
look for dsa you have leetcode and all so dsa is not very boring
if u are very new to programming i would say dsa
Right now I'm in a transition phase between college and exams so I'm self learning python. I have a firm grasp on the basics ... But I'm not able to decide between OOP and DSA .
Aim: I want to become an AI/ML engineer
Yo
DSA sounds better for ur aim i guess.
Yo 1.616255e-35
doesnt matter what is your aim
i think dsa is important because in the process of practicing it you develop this ability to think the logic and code
again i am very new to all this too
but i did develop some programming habbit because of dsa
🤘
Thank you guys!!
You're welcome 1.616255e-35
in what units
Where you from??
i am from Terra
Anyone from Gaia?
Neh
why is …e29 and things used rather than scientific notation with … * 10^whatever
5e6 is an established convention, it's not something Python invented. And it's shorter than 5 * 10**6
https://en.wikipedia.org/wiki/Scientific_notation#E_notation
<...> Because superscript exponents <...> can be inconvenient to display or type, the letter "E" or "e" (for "exponent") is often used to represent "times ten raised to the power of"
I know it’s not a Python thing, I’ve just never understood why that notation is used when *10^x is more easily understood (end much better looking)
also m/s rather than ms^-1
hey guys
ms^-1 just looks better, and ms^-2 is better than m/s/s
can anyone help me in the #1035199133436354600
Once you know what 5e6 means, is it really harder to understand than 5 * 10**6?
ms⁻¹ is not better, and nobody would actually do m/s/s. instead m/s² or, if one must avoid exponents in any form, something like (m/s)/s
m/s oh hey look a single line
yeah, I was thinking that m/s in text refers to the vertical m/s
of course, if you really want to be chaotic, eschew metres per second for metre-hertz
I feel it is justified to be annoyed by kph meaning kilometres per hour
I can understand it on road signs, but why not km/h elsewhere
Or kmh
I was thinking of units like m^3 kg^-1 s^-2 which would be really awkward to write with a slash
This is what made my love of exponents so concrete
How many lines do professional swe’s write/modify in a work day?
m^3 / (kg s^2) what horrors
maybe 2
Specifically this unit
For some reason our road signs never mention the unit of speed. Only for distance, weight, and frequency
if the sign is for a speed limit of 60 km/h, it just says 60
are they many (any?) countries for the which the speed may be in more than a single unit
y'all should standardize
sometimes -5
Imagine that you're driving from the US to Canada. In the US road signs say miles per hour, in Canada they say km/h.
so that you don't forget which unit is used
near the border, twould be wise to have units, sure
I think there might be a sign right after crossing the border that says something like "Alert, the road signs are now in mp/h"
probably
apparently different parts of Canada have slightly different signs, according to Wikipedia
um
maybe it's a good idea to move to #ot1-perplexing-regexing
hello
I guess UK? Metric except for distances and speeds on roads
(And beer)
Yes, and frequently they have both measures on them
Out here speed limits in km/hr are encircled with a big red circle. (They're all km/hr here anyway now, but the convention dates from the cutover.)
... like the 60 sign shown here: https://en.wikipedia.org/wiki/Kilometres_per_hour
but we have a bare 60, no units.
Unfortunate embed. There's another pic on the page itself a little lower down.
rip to eso i guess
news organisations such as Reuters and The Economist require "kph"
hello
why does this speedometer start at 20...
unstoppable car
welcome back 🙂
my pc broke
😮
is 64 gb ram nvidia quadron k2200 and intel (r) xeon (r) e5 1650 v4 good
thats my new setup
You don't really need powerful specs for python
decent programming pc
im js asking
Where did you get that CPU?
well... something new will be in that bucket
old cpus are still not slow, but they tend to use more power than somthing produced now
actually, it's only 3 years older than my CPU. i forgot how time works
64GB of RAM is good though
im getting a new desktop later this year
This year has been kind of lack luster in terms of hardware releases
We'll see
The ones that recently released are a couple % faster at like double the price
ill have to wait and see and compare it with the threadripper pro
how did this chanel specifically made for PYTHON TALKS turn into the channel where we talk about CPUs 💀
you can run one python instance pr core you have
so for some python workloads, having a good cpu makes sense if you have to use python
pip install poop
Perhaps it would be inaccurate at such speeds. Does leave some extra angular room for the graduations of the rest of the scale.
yes, old european cars used cables and tension to show speed, and the machinics involved makes it most accurate in the middle of this distrobution
so more in regular driving speed
some good old german engineering 😄
im personally dont like that we measure distance/time for cars
i find the inverse beeing much better
A CPU that isn't from 11 years ago
The ram is a huge overkill and prob only new ish part in the pc
64 gigs of ram is 64 gigs of ram
What is your goal with that pc?
If your webserver is python you can do a lot with that
yeah, ram is never overkill, it depends what you use it for
Hm.
How much ram does my webserver currently use...
about 40 meg per instance, of which there's 6, for my api
only 40
and 9.6mb for my rust engine I serve my blog from
uvicorn?
yes
its such a great written piece of software

