/tech/ - Technology★

Good for her. Why are you spreading it?

>>879963 (OP)

>Why is shit like this getting circulated?

I really don't care as long as it keeps her busy in front of a computer instead of screaming "Allahu Akbar!" and detonating herself in a crowded marketplace.

>>879963 (OP)

Are you expecting a circlejerk where everyone laughs at her? That's cool and all, but it seems kind of pointless.

>Bourne shell

Bourne shell != bash

>which cd

using which on an obvious shell builtin

>cat /usr/bin/cd

Thinking that this binary is the builtin

>For one, the source for Bash does not have a mirror on GitHub, so I had to browse through a hosted version with a somewhat cumbersome file browser.

She didn't have to do it that way. She could have cloned the official git repository to her computer. Else, she could just browse an unofficial repository of bash on github.

>Usually, [system calls are used] for commands that require a certain amount of privilege.

Completely misses the point of what their purpose is.

>>879993

>using which on an obvious shell builtin

She has yet to discover type!

>Thinking that this binary is the builtin

Ehy, stop the misogyny!!!

Yeah I saw this series posted on hackernews just after a feminist article. It's amazing how one gathered lots of criticism for her superficial analysis and lack of knowledge, and the other had feminists defending the narrative 24/7.

>>879971

Why do you care if she does that?

Her Linkedin is even funnier

https://www.linkedin.com/in/safiaabdalla

>Software Engineer

>CEO

>Data Analyst

>Lead Support Specialist

>I was catching up with my mom on the phone and my baby brother yelled "I love memes!" at the phone really loudly.

<https://twitter.com/captainsafia/status/971945025952198656

I bet 90% of posters here didn't know 'which' command before this thread smh.

>>880054

Shit. You got me! I only know 'where' but not 'which'.

>>879963 (OP)

>Why is shit like this getting circulated?

you fucking retard

>>880054

>smh

This is the grown-up version of the kid that goes "I totally know how to use a computer. Look, I know how to get to Windows Services and I know what most of them do!" except those were usually boys. Now Girls are it.

>>880054

I bet 90% of posters here didn't know 'convert' command before this thread smh.

>someone makes an dumb blog post

>but this time it's by a girl, you guys have to see it

thanks for sharing OP

>>879963 (OP)

But it's (you) who is actively circulating it right now.

>>880063

>>880098

>>880092

Context is everything. Had OP put this on HN, it would have been contributing to her popularity. Here on 8ch, I'd say it has the opposite effect.

Hopefully this stays up so she can look back in ten years and realize how new she was.

The "current directory" or "working directory" is a way for brainlets to imagine themselves "walking" "up and down" a tree because they can't understand that the whole thing is there at once and they're not really moving anywhere. I don't like those kinds of arbitrary limitations.

>>880182

It's so every process doesn't need every path to be absolute, and so you don't have to type the full path of a file every time, dumbass.

>>880148

>implpying 8chan has any effect

>>879963 (OP)

ARCHIVE

https://archive.fo/gFWmz

DO NOT GIVE THIS POO VIEWS

>muh dik tho

>>880223

>putting your dick in poo in the loo

>>880223

>she explains literally nothing

>>880046

i wonder if she's even legal of if her mom was an h1b overstay

This thing is kind of intriguing because it seems to be written by someone with decent general basic knowledge and sense but absolutely no knowledge about anything specific related to Unix. There's a mismatch between using FreeBSD and reading C source code, and not knowing about chdir.

My guess is that it's trying to show the general process of exploring how something works and applying it to the most basic possible example.

>>880044

>all those overlapping "career" dates

Also,

>voulenteer at "ChinkTech"

Have we been trolled?

>>880234

>purchase business license for $100

<you are now ceo

>shill "business" to every "muh womens/minority" business support organization and apply for all of the government grants

>...

>...

profit!

it's 2018 you can change your race and gender right now and do the same.

>>880234

That says ChickTech you dingus

>>880200

Hang in there, buddy. Try to find jobs in the state and become a parasite like me. In my country, state employees literally cannot be fired, have better salary than private employee cucks and work like 20-30 hours per week. This is often the way out. Apply your autism to the task.

I liked the article. I'm vaguely familiar with Linux but most of the details were new to me.

>>880231

This is insightful, but this "mismatch" is actually pretty common. I studied math and eventually got into machine learning. I've done a lot of programming in a wide variety of domains, but only picked up bits and pieces of Linux knowledge along the way. I'm still a very basic Linux user.

As someone who now (in spite of my patchy knowledge of Linux) works on the boundary of machine learning and systems programming, I see both sides. There are ML experts who don't know what a process is. And there are systems programmers who make basic mistakes and math and stats. Both would be cause to get annoyed if you were that kind of person. But if you give people the benefit of the doubt and don't judge them for not knowing what you don't know, most people can understand their mistakes and learn as they go.

>>880386

>civil service

Sure beats NEETbux, is more dignifying and has to be more useful than whatever your politicians are doing.

>>880360

>>880148

>>880044

>girl makes blog to share and document her learning experiences (as everyone in any sort of information field should do)

>blogpost manages to make front page of hn (as they usually do from time to time)

>proceeds to make halfchan/pol/ /g/ tier outrage post instead of actually doing things

please see here before you post here again >>860932

>>880200

>autistic, iq not high enough

Why isn't your IQ high enough? Are you sure this is the case, or are you just suffering from a perceived inferiority complex. This afflicts many smart people, consider the case of a very smart math prodigy joining an elite school. Unlike in every other instance of this man's life, now everyone around him is competent, and likely knows more than he does. He is goes from genius to ordinary quite quickly. Of course what he doesn't realize is that everyone else around him has the same feeling.

For the more you know, the more you are aware of your own ignorance. This allows self doubt to creep in as you suspect your peers might be better versions of you.

>>880396

Hello SAIDF!

>>880398

This isn't really true, most intelligent people never gave their intelligence a second thought (look at how Feynman excelled at the Putnam test without even trying). Although, sure, you could be putting yourself down needlessly - and that is a bad mindset to have.

>>880403

>This isn't true

>Most...

>Cites one physics prodigy

I have encountered many smart people who tell me they feel like frauds in departments they are experts in. Also if you read FLoP, he's quite humble about Mathematics.

>Consequently, the proof may look deceptively simple, when in fact, the author might have worked for hours trying different ways of calculating the same thing until he has found the neatest way, so as to be able to show that it can be shown in the shortest amount of time! The thing to be remembered, when seeing a proof, is not the proof itself, but rather that it can be shown that such and such is true. Of course, if the proof involves some mathematical procedures or “tricks” that one has not seen before, attention should be given not to the trick exactly, but to the mathematical idea involved.

>>880410

this can't be real

I don't even know who this girl is nor why I should care about her blog.

>Why is shit like this getting circulated?

>>879965

>>880423

Here I am at the bottom of this shitty thread and I still haven't seen her tits. What a fucking waste.

>>880421

That's a shame, because having an IQ is quite fun. The hardest part is deciding what you want to spend your free time on next. Math, physics, electronics, astronomy etc. There's just not enough time in the day.

>>880457

You can stop posting about this sand nigger now, no one cares mate.

>>880457

She might not know how `cd` works, but mate, you dont even know how averages work

>>880485

11,100 tweets

4 solid years

---

2775 /year

7.6 /day

...

>>880455

And then you learn about the jews. Being intelligent _and_ mentally strong isn't quite the fun in decadent societies.

this thread would be more appropriate on /cow/

and raid her afterwards

>>880468

Listen, safia, if you want us to stop posting about you, you should just identify yourself.

>>880011

>why do you care if she kills other people

your brain on cultural marxism

>>880449

>>880450

>>880451

Absolute gross...this doesn't belong here

>>880557

Are there more pictures like this?

>>880689

Don't be a faggot anon. Reverse image search it, go to your choice of imagedumping site, click on the tag for the author.

>>880691

I did search. Turned up literally nothing.

>The Unix kernel executes the chdir call and does its own low-level thing.

>I could dive in a little bit more into how #4 works, but let’s be honest, I’ve already read too much code at this point, and my eyes are starting to hurt.

>I hope reading this was illuminating for you!

>>880694

>reverse search your image

>0 results

>>880461

>nickers

niggers

>>881095

probably because iqdb is for 2d and not 3dpd.

>>881102

I'm not a retard, I used google, yandex and tinyeye, along with multiple others.

Before reading this, it never occurred to me that you can't implement 'cd' as a program. It has to be a shell intrinsic.

I never really though of how or who called chdir(). It's funny, though, that her conclusion is pretty much what I thought of cd:

1) resolve directory name

2) call chdir()

3) kernel magic

4) ...

5) profit

>>881106

I'm curious on why you thought that you could implement as a regular application. Can you explain how you thought how a program like that would work?

>>881107

I had just never thought about it. It never occurred to me to want to know how it worked.

As far as I can tell, chdir() literally just changes the field pwd in the program's fs_struct of its task_struct. What that means is beyond me (as far as the kernel goes; I know what its effects are)

>>881110

I was just asking as it seems natural to assume that the shell has to keep track of what directory it is in somehow, so it wouldn't make sense to use another program to edit the shell's state.

>>881112

I've never really had the desire to write a shell. You?

>>881113

No?

>>881114

Was it just to learn, or did it have a killer feature that you wanted? There are so many shells that I have had no desire for either.

So many salty faggots in this thread. This is what young tech people are now. My own younger cousin (male) is exactly like this.

Get on with it, old people.

>>881115

I never said I wrote a shell as fish serves me very well. That's not say I haven't exercised my rights of modifying it before.

I don't have cd or chdir in /usr/bin. And reading the bash source all it does is call chdir. Which is totally expected considering how processes work.

So idk what this was supposed to teach me.

>>881118

you have to go back Safia

I figured out how to listen to CD's when I was five. Really not hard to put a disc in a tray.

>>881164

Imagine being the kid who didn't know why his PC didn't play CD audio because older CD drives have an extra connector for it that links to the sound card. Really sucked for games that used Redbook audio for music such as LBA2.

>>881166

I remember unplugging the printer to plug in one of those snes-like gamepads with the purple d-pad.

>>880541

She does not have to make 4-5 tweets a day. That number is an average. Maybe she makes 20 in one day and a couple over the next few. Maybe she didn't use it properly for a year and makes 10 a day. Don't form opinions with such lazy foundations

>>881200

are you serious? what opinions should be formed? oh hey there's 11 thousand tweets posted here that's perfectly normal and acceptable it's 2018 we should just willingly post 11 thousand tweets about our daily lives over any length of time into a company that exists for the sole purpose of selling that data to anyone who wants it along with real names and phone numbers and email addresses.

>>881183

I confess I didn't know it was an actual anime. I thought it was just a series of pictures someone drew.

>>881167

I thought most game controllers used the game port. That was also used as a MIDI interface, for a time I used to have a digital keyboard that needed a PC to output anything, before switching to a dedicated box.

>>881200

The mind of a defender of a feminist, folks.

>>881228

Python is good.

Tumblr is garbage, but it's garbage that's used by a lot of people.

>>881229

Python is good for some things but it's still a hipster language. She uses Java and Javascript too but she knows nothing else.

https://cloudup.com/cqozUYp889k - her CV

>>881231

Python may be a hipster language, but it's also a data science language, and a Google language, and an educational language, and a lot of other things. It's used all over the place.

>>881239

>has her own site

>has her own mail server

>worked as intern while attending university

>probably has more riced vim than average /tech/ poster who uses VScode

>>881248

>bought a domain name and typed it into the tumblr dashboard

>Figured out how smtp works

Why are you defending her?>>881248

>>881236

And it's one of the slowest used languages.

>>881276

Yes, and? For some of its use cases you might only need to run a given program a few times, or even once, depending on what it's for. Python is great for throwing such things together quickly, that you save much more on programmer time than you might gain on fastest execution. Think writing a shell script instead of doing it in C.

>>881276

For most of my Python code the bottleneck is in the network, or in the filesystem, or in the natively implemented parts of the standard library, or in numpy, or in some other library, or in some non-Python process, or unimportant because no heavy computation is needed in the first place.

Python's speed is not a big deal for 90% of the things Python is actually used for. The tradeoffs it makes don't make sense for all languages, but there is a valid niche for languages that make them the way Python does.

>>881282

The thing is that it's slow without reasons other than the foundation not funding/working on pypy.

I honestly have only three problem with Python:

1) A scripting language shouldn't be used to make more than scripts

2) Dynamic typing is shit for brainlets

3) A lot of what python does can be done easily with POSIX sh/awk, but the faggots can't live without a stdlib the size of a billion whales

>>881239

>tfw she has more experience than me and it's more extensive

what the fuck…

>>881314

>1) A scripting language shouldn't be used to make more than scripts

"Scripting language" and "script" are defined vaguely enough to make this either meaningless or tautological.

>2) Dynamic typing is shit for brainlets

It's a blessing for a lot of tasks, especially combined with duck typing. Easy things aren't automatically bad, although they're likely to be bad for some things - but everything is bad for at least some things.

>3) A lot of what python does can be done easily with POSIX sh/awk, but the faggots can't live without a stdlib the size of a billion whales

I'm actually trying to cut down on POSIX sh because it's a really, really, really bad language. It's fucking awful. It's worse than PHP. And because it's just barely good enough to fill its niche there are no better languages to take its place. All its strong points are because of the things it tries to do, not because of the way it does them.

I use POSIX sh for very simple text processing and for very thin wrappers around other software, but I stopped using it for tasks beyond that.

Python can do just about everything sh does, while sh is garbage at most of the tasks Python does. Not everyone wants to use two languages, and writing correct (not just working) sh is unreasonably tedious and tricky. Python is already installed on every single system I use.

>>881352

Have you heard of Oil? Curious what your thoughts on it would be.

https://www.oilshell.org/blog/2018/01/28.html

>>881358

I've heard of it, but I haven't tried it.

This code example bothers me:

proc make_hdb {
  # Some distros don't put /sbin:/usr/sbin in the $PATH for non-root users.
  if test -z $[which mke2fs] || test -z $[which tune2fs] {
    export PATH = "/sbin:/usr/sbin:$PATH"
  }

  truncate -s $(HDBMEGS)m $HDB      &&
  mke2fs -q -b 1024 -F $HDB -i 4096 &&
  tune2fs -j -c 0 -i 0 $HDB

  test $Status -ne 0 && exit 1
}

$(), ${} and $[] are being juggled around. Bash's ${} is now $[], bash's ${} is now $(), and (not visible in that example) bash's $[]/$(()) is now also $(). Why would you do that? You shouldn't keep bash's syntax as it is, but shuffling around everything so you still have the same syntax but now all of it means something different that already had existing syntax seems unnecessarily confusing.

I do use fish a lot. Mostly interactively, but I've been writing more and more advanced shell functions. Here's the (untested) fish function equivalent of that oil process.

function make_hdb
  # Some distros don't put /sbin:/usr/sbin in the $PATH for non-root users.
  if test -z (which mke2fs); or test -z (which tune2fs)
    set -x PATH /sbin /usr/sbin $PATH
  end

  truncate -s {$HDBMEGS}m $HDB
  and mke2fs -q -b 1024 -F $HDB -i 4096
  and tune2fs -j -c 0 -i 0 $HDB

  test $status -ne 0; and return 1
end

A lot of syntax is implemented as (special) builtins. Instead of '&&' you use 'and', instead of '||' you use 'or', instead of brackets you use 'end'. This is weird at first, but it makes the language a lot more consistent and easy to deal with.

For example, the 'function' builtin takes options, like a regular command. If you want to run a function every time a command finishes, just define it with 'function -e fish_postexec'. If you want a function to inherit completions from the foo command, define it with 'function -w foo'. If you want to know which options 'function' takes, run 'function --help' - you can do that with almost every built-in. It makes the syntax a lot easier to explore and remember and recognize.

Oil's export looks like it's still falling into the trap of imitating syntax within a command's parameters. Fish's "set" very straightforwardly functions normal command syntax.

Fish doesn't have arithmetic syntax. There's a POSIX tool for that. Bash's $((x + 1)) becomes $(x + 1) in oil, but in fish it's just (expr $x + 1). That's much nicer. Why make your own arithmetic sublanguage if there's already one installed?

Aside from weird pseudo-builtins and external programs, fish also uses a lot of "proper" functions and built-ins. There's no $#, just use the "count" command, which outputs its number of arguments. There's an 'alias' function, but it's just a function that's loaded by default, and if you run 'type alias' you get a syntax-highlighted dump of its definition - no tricky business, all of it things you could have written and could in fact copy to your own alias implementation with extended syntax.

Fish's arrays are really nice. To load the lines of foo.txt into an array called $bar, just run 'set bar (cat foo.txt)'. A string with newlines automatically becomes an array. This works well with line-oriented tools (half of Unix).

An array behaves approximately like "$@" for most purposes, and you can also set multiple values by giving set multiple arguments, like in the example.

$PATH is special-cased to be translated on the fly. For the common case of adding extra entries to it there's an additional variable, $fish_user_paths, which is useful for avoiding either adding the same directory to the PATH every time you launch a new shell or removing directories that were in the pre-existing PATH.

Oil does look like it's much more suitable for complex programs than fish. If it gets packaged by the distro I use I'll definitely give it a try, until then installing it on all my systems is enough of a bother to stop me.

>>881352

Comparing sh and Python is exactly what you shouldn't do. One is an executable-gluing, data pipeline oriented language while the other is a behemoth between the programming and scripting language.

If you don't like it, you could use something like ksh/bash and still be as or more portable than Python while having your goodies. And awk still stands; you can do much with this small forgotten language.

>>881387

ksh and bash are backward compatible with POSIX sh, which automatically makes them shit. fish is almost bearable.

I think Python is more portable than ksh, bash or even POSIX sh, at least for the things I'm likely to do. sh is inextricably tied to Unix, if you want to run them on Windows (I don't use Windows, but it's entirely possible that I might have to some day) you need to install an entire Unix environment. How is that portable?

Even if you stick with Unix it's not very portable. At my current job where everybody else uses MacOS we use maybe a hundred lines of shell in total, but I still had to fix an issue caused by GNU sed's -i option behaving slightly differently from MacOS's sed. A hundred lines of code, two platforms, and there were still portability issues!

I honestly have no idea why you would call sh portable.

I don't really know AWK, but I'd like to. It seems nice.

The point I was trying to make is that people will always try to use the tool they know for tasks that are better served by other tools, but using Python for sh-ish tasks is much better than using sh and AWK for Python-ish tasks.

>>881391

>using Python for sh-ish tasks is much better than using sh and AWK for Python-ish tasks.

this part is wright in the benis. 100% true.

>tfw nobody gives a shit that I know how cd works since im not a girl

>>881391

>confusing sh and the core/text/findutils

Are you this retarded? POSIX also specifies sed.

>It's GNU's fault if Apple doesn't want to use anything GPL3

Protip: it isn't

>I honestly have no idea why you would call sh portable.

Because the most widely used OS standard makes it mandatory, along with awk and C. Read https://www.python.org/dev/peps/pep-0011/ and compare with POSIX.

Honestly, if Perl didn't have a shit syntax that makes C++ beautiful, Python would have no reason to exist.

>>881391

Not the other anon but:

>Reeee POSIX is not portable

>proceeds to describe sed -i a non posix option.

Come on man, I do agree that PoSix sucks but complaining about incompatibility between non-standard options is a bit silly.

>awk

Yes you should learn awk its pretty good.

>>881415

>>confusing sh and the core/text/findutils

sh isn't built to be used without the standard Unix utilities. Unix is, roughly speaking, sh's standard library. Pure sh without external programs is a fun game but rarely useful.

>Are you this retarded? POSIX also specifies sed.

What's your point?

>>It's GNU's fault if Apple doesn't want to use anything GPL3

>Protip: it isn't

I never said that. As far as I know they've never used GNU sed, so it isn't even relevant to anything I said. I'm on GNU's side when it comes to that particular issue.

All I said was that the typical implementations of sed installed on MacOS and GNU/Linux are incompatible when it comes to a certain basic function. That much is certainly true. I don't know whose fault it is, I don't care whose fault it is, I don't know if it's anyone's fault in the first place, but I do know it's a pain in the ass.

>Because the most widely used OS standard makes it mandatory, along with awk and C. Read https://www.python.org/dev/peps/pep-0011/ and compare with POSIX.

So it's portable to operating systems that follow that standard. That's excellent, but not all operating systems follow that standard.

I'm much more likely to start using Windows, which doesn't follow POSIX, than to start using any operating system that CPython doesn't support. What I care about when it comes to portablity is running my code on the systems I use.

>>881417

POSIX is pretty portable, but nobody sticks to it, because it sucks. POSIX isn't terribly relevant for deciding whether sh/AWK is portable, what's relevant is the set of the functionality that people need and use in practice. And that set isn't very portable.

With "POSIX sh" I meant just the shell, not the rest of POSIX.

>>881422

>With "POSIX sh" I meant just the shell, not the rest of POSIX.

Sed isn't sh though so how is that relevant to posix sh?

>All I said was that the typical implementations of sed installed on MacOS and GNU/Linux are incompatible when it comes to a certain basic function.

Its not a basic function; though it is a useless one. Whats wrong with sed 's:faggot:nigger:' <file >file.new && mv file.new file

I can't imagine you can find a version of sed that would find this unacceptable. You are already executing it in a script so whats the difference if you need an extra command after sed.

>>881422

>What's your point?

My point is don't cry about portability when using non portable options. While you're right that POSIX is lacking in certain places (to the extreme in some; handling \n in filenames, for example), writing POSIX compatible C/sh isn't hard at all. Especially if you're ready to complement sh with some C wrapper around extremely important functions (see the realpath/readlink -f situation) and emulate some stuff with some sh functions.

For example, you can have stuff like

# Portable echo, without any option
echop()
{
    printf '%s\n' "$*"
}

# Portable pgrep
pgrep()
{
    _psout=$(ps -A -opid=,args= | awk -v regex="$1" \
        '($0 ~ "^ *[0-9]+ " regex "$") {print $1}')
    if [ "$_psout" ]
    then
        echop "$_psout"
    else
        return 1
    fi
}

# List the absolute path of all files in $1. The remaining arguments are passed
# to find.
listfiles()
{
    _dir=$1
    shift
    find "$_dir" \( ! -name "$(basename -- "$_dir")" -prune \) "$@"
}

match()
{
    echop "$1" | grep -Eqx -- "$2"
}

# Portable head -n-val
headneg()
{
    ! match "$1" '[[:digit:]]+' && return 1
    awk -varg=$1 '{if(NR > arg) print buf[NR % arg]; buf[NR % arg] = $0}'
}

to make your life easier.

>>879963 (OP)

>another person looks at the code of sudo and sees nothing wrong

this triggers me

>>881428

>Sed isn't sh though so how is that relevant to posix sh?

sh is meant to be used with external tools like sed. How many people use sh without external tools?

I tried reimplementing them in pure POSIX sh once, and I got as far as cp before I quit, but it wasn't pretty.

>Whats wrong with sed 's:faggot:nigger:' <file >file.new && mv file.new file

It takes more effort to write. But I didn't write the non-portable code, so it doesn't even matter what I would do.

>>881432

Why call sh more portable than Python at this point? You're writing a custom wrapper to simulate echo, how is that supposed to show that sh is portable?

People are going to do non-portable things all the time in it, whether I like it or not. I know it's possible to write portable sh, but that doesn't make it a good example of portability.

>>881239

>Java, Python, Javascript, Node.js, Go, CSS, SASS, HTML

>React, Express, Redux, RxJS

why am i not surprised at all.

>>881326

it's easy to get these jobs when you've got multiple diversity boxes checked, notice she got most of this before the degree.

>>881248

>own mail server

more like told gmail to use her domain

>>881313

Whoa slow down, it's almost like you thought you were talking to a real programmer. This is a LARP board

>>881564

Are you seriously implying real programmers use python? Thats the bullshit that every highschool kid learns because they can get a gui to popup in 3 lines of code.

>>881565

Oh fuck my bad

<hides power level

You're right, if its not in C it gets the pee.

>>881566

Knowing C is a sign of not being a fucking retarded that does not know how to use a computer. I know plenty of actual real python programmers, but despite that almost every person I have met that "knows python / javascript" is a fucking retard that does not know trivial shit.

>>881568

I know, right? It's like they don't even know that C is the best programming language for every situation. If it aint C, get away from me!

>>881577

Okay fuckwit I literally just explained how that's not the point but you seem to be on a role with your thing.

>>881584

Hey man, I'm agree with your premise that if somebody isn't using C for everything, they must not know it.

>>881592

Where did anyone say that?

>>881577

>C is the best programming language for every situation

correct

>>881595

>>881565

>>881565

Real programmers use Python.

Unskilled beginner programmers use Python, too. I think that's a sign it's succeeding in its niche.

It doesn't mean that "real programmers" shouldn't use Python.

>>881657

it's also why portage takes 5 minutes to calculate dependencies

>>881663

Of course it's a bad language for some jobs. What's your point?

>>881667

The point is that it's a bad language for ALL jobs. I've used python a lot, and I used to like it, but it's terrible at everything it does. it's bloated and slow in every way that can be measured.

It's ONLY benefit is ease of use.

>>881672

Compare to C and assembly, which are good at ALL jobs.

>>881657

Beginners use Python because other people tell them it's good for beginners. It's just another trick. Like recommending CLRS, SICP or TAOCP, when most programmers read literally nothing but tutorials. Or academics tricking people into learning category theory.

>>881672

>>881674

You're still free to write everything in C if you want to.

The most important is to enjoy the process, right?

>>881657

>It doesn't mean that "real programmers" shouldn't use Python.

Of course it does. Because now this means that Python language is easy, and as we know, real programmer should not ever use something that is easy. It's the #1 rule of /tech.

Everyone who uses an easy language cannot be a real programmer, they're all wannabe programmers and girls.

>>881674

>assembly in charge of portability

>C in charge of shell scripts

Stop with your hard-on for the One True Programming Language. Such a thing doesn't exist.

>>881672

Is it a bad language for my cron job that spends most of its time waiting for network requests?

Is it a bad language for my data analysis script that uses numpy and would be slower, buggier, have less features, and take more time to develop if I wrote it in C?

Is it a bad language for the script I needed to use once that took two minutes to write and five seconds to run?

>>881682

Is it C?

If it's not, it's bad, because it's too easy to write, even soyboys could use it.

That's what you're going to get on /tech, a board full of LARPers.

It's kind of pointless to argue with them.

>>881568

Problem with languages like python, javascript, etc. that are used as "teaching language" is not that they are somehow unreasonably shitty languages. They are not perfect ether, but every programming language has some shitty parts.

Problem with programming is that it has some rough requirements if you want to be competent programmer, like reasonably high intelligence, logic and math* skills, genuine interest in programming, will to learn zillion things, etc.

Thing is, we do not know how to measure those things that make good programmer. Best way to find out if someone could be a good programmer that we came up with is to expose them to programming and see if it "clicks" with them. Unfortunately that creates a lot of bad programmers who use those "teaching languages".

Another problem with programming is that it has to compete for those good people with other fields like engineering, science and math that require roughly same set of traits. That leaves only some small percentage of people as competent programmers, which provides competent programmers with a lot of leverage to demand good wages, better work environments and whatnot which then create a lot of reasons for companies to funnel bad people into programming.

I am just going to end post here because I do not want it to be too long. There is a lot that could be written about relations between bad programmers, frameworks, various programming philosophies, adoption rates of different technologies and wage suppression.

*depends on what you work on, some problems do not require a lot of math

>>881695

You forgot the part where you describe the problem with the language itself.

>>881695

Your point seems to be, roughly, that

- it's worse for someone to be a bad programmer than for them to not be a programmer at all

- there's no way to make good programmers that doesn't also produce bad programmers

That's an interesting idea, but I don't see how it relates to this discussion.

>>881682

>Is it a bad language for my cron job that spends most of its time waiting for network requests?

Yes

>Is it a bad language for my data analysis script that uses numpy and would be slower, buggier, have less features, and take more time to develop if I wrote it in C?

Yes

>Is it a bad language for the script I needed to use once that took two minutes to write and five seconds to run?

Yes

>>881721

What would be a good language?

>>881722

C

/tech/ has come full circle. We no longer have to larp being useful to society, now we just larp about being a larper.

>>881735

>all these white knights in this thread wanting to smell the curry

Sorry for the unclear writing, I am not native English speaker so my writings sometimes end up being incoherent wall of text.

>>881696

>You forgot the part where you describe the problem with the language itself.

I am not arguing that languages do not have bad parts and various problems. I am arguing that some languages get bad reputation just because bad programmers use them to create low quality software.

Take Javascript for example, some people hate it because webdevs use it to make 50MB 'webapp' just to display some static text. If Eich implemented Scheme (as he orignally intended to) instead JS there would be 50MB Scheme 'webapps' to display static text. It does not matter which language they are using, they would still make terrible things.

There is no reason to shy away from using some language just because there is community of people using it badly. IMO debating about programming languages without specifying constraints and intended usage is, in most cases, nothing more than language popularity contest.

>>881699

>That's an interesting idea, but I don't see how it relates to this discussion.

I was trying to make other points as well:

-using whatever language does not make you a bad programmer per se

-there is much higher chance to encounter bad programmer who uses one of "teaching languages" than other languages

-languages with community of bad programmers get labeled as bad languages

>>881681

>assembly in charge of portability

Just rewrite it for every platform, idiot. What are you, a webdev?

>C in charge of shell scripts

Lol why do you need that? Just recompile every time.

>>881675

Hey now, I use category theory every day. You only have to remember a few simple definitions.

>>881721

>>Is it a bad language for my cron job that spends most of its time waiting for network requests?

Yes, dynamic types lead to many trivial errors.

>Is it a bad language for my data analysis script that uses numpy and would be slower, buggier, have less features, and take more time to develop if I wrote it in C?

It is both a bad language for data analysis, and it would be a shitty idea to write in c.

>Is it a bad language for the script I needed to use once that took two minutes to write and five seconds to run?

If it took 2 minutes to write then its so trivial it does not matter.

>>881685

>>881682

>>881681

>>881674

>butthurt python "developer" pissed that he is being called out for having "learned" a programming language.

If you faggots actually knew a programming language this would not be so bad. But its always these people that "know" python, but don't know what context managers, or generators are.

>>881861

It's reasonable for many Python programmers not to know what they are, because they don't need to know. Anything worth doing well is worth half-assing.

It's better for someone to know some basic Python than for someone not to know how to program at all, and for a lot of those people it's unreasonable to expect them to become experienced knowledgeable programmers, because they use programming as an occasional tool to help with other activities.

Luckily, Python is built in such a way that you don't need to know what context managers and generators are in order to use them. People can start using the "with open(...) as f:" construct when they're just starting to learn Python without knowing how it works and that it's part of a more general feature, and they'll still benefit from it. People can loop through generators without knowing that they're not actually containers.

I do really know Python, even if I don't expect all Python programmers to. You can quiz me if you want. I only made one of those posts, though, so I can't speak for the others.

>>881871

Python has many constructs that make it harmful when interacting with especially jr developers. It's almost impossible to actually hide implementation details to preserve invariants for an object. The type system dynamic which means that code the jr developer wrote won't fuck up until you are actually reproducing the specific case it breaks in. Pythons copy semantics can particularly be confusing to newbs as with these types of languages. They think they have made a copy of an object but actually just threaded the same instance into 50 parts of their program fucking up in very strange ways. There are complaints people have like "muh no curly braces" but that's a trivial reason to hate a language. The only good thing I like about python as a language is the fact that the library ecosystem is very easy to use without much knowledge, besides that though its horrid. Want a basic GUI window? 10 lines of code, etc.

>>881873

>but that's a trivial reason

A trivial thing that I HATE about python is how it complains about improper white space. Sometimes I just want to comment out a loop / conditional without having to reformat the rest of the code.

>>880360

>wrist-cutting direction is on point

>gun aimed temple instead of roof of mouth (brainstem)

this really gets my noggin joggin

>>881882

they goyim can't be trusted to write readable code for later copying and profit so the language will enforce it for them.

>>881748

She dug into how cd worked until she got into the kernel, then realized that she was out of her league and quit. If you want to say that cd calling chdir is "knowing" how cd works, fine. Honestly, I dug into the kernel code and still can't tell you what the fuck is going on other than >>881110 . Knowing that cd calls chdir is like knowing that gasoline makes a car go: it is a very superficial understanding of what is going on. The only important things she demonstrated was the methods by which she dug into the problem: however, I am disheartened at her balking at reading C and balking at using anything but GitHub. Savannah isn't that bad.

>>881984

Setting pwd is pretty much all that needs to happen. If you are looking for what is happening with pwd, there's only a couple things. Stuff that you would imagine like the getcwd syscall, /proc/ entry, etc. The most important place is fs/namei.c. This is where the logic for resolving paths are.

In an abstract point of view cd's only point is to change how path's are resolved so you, and other programs that you are running, do not have to use absolute paths.

>>879963 (OP)

>You'll need to be on Tumblr to see this particular blog

yea nah

>>882022

SHE SHUT IT DOWN

SHE LOGIN WALLED THE BLOG

it was wide open before.

>>880076

>people only know the commands they use

I learned Unix as a teenager by having somebody teach me the apropos command, man, and ls and being told to look shit up in the manual. It took me years to bother to look up some commands.

>>881882

in PyCharm this is absolutely not a problem.

also, you can do a "dummy loop": `for _ in range(0,1):` if you really want to keep indentation as a temporary "hack"

>>881861

>but don't know what context managers, or generators are.

yeah that's pretty stupid to not know generators.

this is one of the several things that make it so great. and it's not even that hard to understand even for girls, I recently taught this concept to one and it mostly worked.

btw you also should have mentioned coroutines (as in asyncio-n-stuff for example)

>>881873

it's always possible to fuck with other code, in almost any language. using things like reflection or raw pointers (optionally through some FFI if the language doesn't expose them), or what else is available. trying to solve non-technical problem with tech is an exercise in futility.

>>881873

cross-platform GUI in general is such a complex beast it doesn't belong to stdlib. Tk doesn't count, you won't make a good GUI with it anyway. quality of third-party libraries is not the language's problem. I remember PyQT having weird quirks related to reference counting, but it's certainly possible to write something else and use it.

>>881860

>If it took 2 minutes to write then its so trivial it does not matter.

it will start to matter if you decide to write it in C.

>>881860

>Yes, dynamic types lead to many trivial errors.

python has optional static typing, and tools which can check them statically. you just have an option to ignore that when it's not needed.

>>882033

>you did it /tech/

did you even doubt us?

if there's anything we shine at, it's the art of trolling.

>>881859

Fine I'll bite anon. What do you use it for?

>>882038

>python has optional static typing,

Dynamic languages with bolted on progressive type systems are absolutely shit to work with especially when the entire ecosystem does not use it. You end up doing dynamic type checks constantly for your supposed statically typed code because of type change barriers.

>it will start to matter if you decide to write it in C.

I agree, trivial programs that can be written in two minutes are a great use for python.

>cross-platform GUI in general is such a complex beast it doesn't belong to stdlib

I was sayyyying that it was very convenient to be able to open up gui quickly.

>it's always possible to fuck with other code, in almost any language

Not equally. Things actually are different.

>using things like reflection

Can be done safely with type systems

> or raw pointers

are bad

>trying to solve non-technical problem with tech is an exercise in futility

If you think having people using raw pointers for everything like c is a non technical problem you are retarded. Its a clear fault of the language that has lead to endless bugs.

>>882477

>pointers without bloat are a fault of the language

kill yourself

>>882492

Nope, but editing raw bytes when you care about a list is retarded.

>>882494

Yeah, but there is no list type in a x86 / x64 processor. That's why you need to use a pointer when using the list. You only have to write a correct implementation of a list and then you don't have to worry about it.

>>882505

>You only have to write a correct implementation of a list and then you don't have to worry about it.

Great, then I implement the list with a pointer, and then never use pointers again (which is what safe languages do) because they are unsafe. In a language like C though its impossible to abstract away the unsafe shit, and every trivial string operation is dangerous.

>>882510

>which is what safe languages do

the performance on these "safe" nu-languages is also terrible, fuck your safety.

>every trivial string operation is dangerous.

you should go to your safe space and hang yourself there rustfag

>>882514

>the performance on these "safe" nu-languages is also terrible, fuck your safety.

You know that you can have memory safety without being dynamically typed python bullshit right? C++ and way more memory safe than C is.

>you should go to your safe space and hang yourself there rustfag

Rust is shit, but at least you don't have to worry about every single piece of your program having the power to fuck with every other part.

>>882516

*C++ is

>881239

Kinda looks like she can't keep a job for long nowadays

>>882516

>C++ and way more memory safe than C is.

it's also way slower.

> don't have to worry about every single piece of your program having the power to fuck with every other part.

i don't want to be artificially limited in what I can or cannot do, this is a dangerous idea and the logical extension of that is being locked into an api that you have no power to deviate from, ie an unrooted android, or windows locking down the app store to the point where only (((approved))) developers can write real code, the rest have to use their shit api wrapped in a security blanket.

>>882525

another more relevant example is chrome and firefox forcing all their extentions to go through their kosher approval process. they would do this at the programming language level and sell api access to lower levels if they could get away with it.

>>882525

>it's also way slower.

You must have literally never looked at any data

>i don't want to be artificially limited in what I can or cannot do

You won't be, you can always drop down to assembly.

>into an api that you have no power to deviate from

When deviating from the API means the program crashes because you are calling a function that does not exist perhaps thats a nice feature.

>kosher approval process.

You are always free to drop down to assembly if you want to

https://rhbtry.blogspot.com/....

>>882477

>Not equally. Things actually are different.

Name me a language where it's not possible to force a leak in any abstraction.

>>882477

>Dynamic languages with bolted on progressive type systems are absolutely shit to work with especially when the entire ecosystem does not use it. You end up doing dynamic type checks constantly for your supposed statically typed code because of type change barriers.

BS. I don't do dynamic type checks and I'm completely fine.

>when the entire ecosystem does not use it

the stdlib has type annotations. the rest will follow (and some already did).

>>882477

>I was sayyyying that it was very convenient to be able to open up gui quickly.

Not needed in any real program.

>>882477

>>882477

>>using things like reflection

>Can be done safely with type systems

>> or raw pointers

>are bad

are you even able to read complex phrases? this part of your reply looks like you're a bot with too few neurons. I only mentioned 2 of the many possible ways to force leaks in abstractions in any language, and the "facts" that they are "bad" or "… can be done safely" is irrelevant

>>882477

>If you think having people using raw pointers for everything like c is a non technical problem you are retarded. Its a clear fault of the language that has lead to endless bugs.

it's not about pointers, it's about all ways to force leaks in abstractions, and pointers are one of the possible ways.

again:

name me a language where it's not possible to force a leak in any abstraction.

then think about why smart programmers don't really want to use it for most programming tasks.

>>882652

>Name me a language where it's not possible to force a leak in any abstraction.

Fractran.

>>882652

>BS. I don't do dynamic type checks and I'm completely fine.

Really? You don't do any dynamic checks? This is bullshit. The entire ecosystem is built around dynamic checking. Even if you could, python is such a shitty language that you have to manually specify your type annotations instead of a compiler automatically doing it for you.

>Not needed in any real program.

O we are talking about real programs now? I thought we were talking about python the toy scripting language.

> leaks in abstractions in any language,

>in any language

You chose a shitty language and then when when its pointed out that the language is shitty you come back with "hurr durr every language can have errors in it". No shit.

> it's about all ways to force leaks in abstractions

And languages are absolutely not equal. Some languages make it much easier than others.

>name me a language where it's not possible to force a leak in any abstraction.

Any abstraction is implemented. Any abstraction not an exact description of physical reality will "leak". Languages are not equal when it comes to leakage.

>>882684

who's using it in production?

>>882686

>you have to manually specify your type annotations instead of a compiler automatically doing it for you.

only when I feel like it's needed.

>>882686

>I thought we were talking about python the toy scripting language.

LOL

this is so stupid, and so easy to debunk by a simple fact checking, I don't feel I need to answer anything below as it probably won't be understood.

>>882769

>fractran

>who's using it in production?

okay, I looked it up, and immediately found that it's an esoteric language. so, not a worthy example even if that claim is true.

>>882769

Oh, so you actually are still writing dynamically typed shit code then. You made it sound like you were letting a static system do the work for you earlier.

>>882806

I'm pretty sure he's being deliberately obtuse. The false dichotomy between leaky and atomic abstractions was a big hint.

>>882721

I love how she just throws public speaking in with cleaning and listening to music

>>880616

This.

>>879963 (OP)

> no archive

> no copied text

> being this transparent with your shilling

tits or gtfo

>>885258

> "Are you sure?" he pleaded.

> Asking someone if they are sure they don't want rewarded for their efforts is now "pleading"

> The shitlord had the audacity to give her "unnecessary cognitive load" and distract her with thoughts of being asked out on a date

>>885328

not if they find the guy attractive or want to fuck their way to the next promotion. it ultimately doesn't matter though. in the modern era, do not give any female at work more attention then is absolutely necessary. do not even be in the same room alone. attempting to get laid at work will never end well, even if you succeed. these men she references are lucky they escaped with some awkwardness, as she get's older and uglier, she will become more desperate and the victim card will be used more often and in a much stronger way.

remember she doesn't make money by writing code. she makes money by being a non-white female who talks about being a non-white female and writing code. being a victim is her career.

>>885258

she also unlocked her shitty blog / advertisement, and deleted the entry in OP.

>>879963 (OP)

>>885391

Do diversity hires mask and obscure real job loss due to automation? After all, inherently useless hires might just as well not count towards job statics, which would mean if you remove all diversity hires from job statistics then the statistics might show declining employment levels.

Why should anons care about her?

This thread is just gonna add even more fire to the 'bad programmer guys you just jealous i can code' bullshit if they discover this thread.

>>885419

she's a poster child for the anti-male anti-white anti-straight pro-left sentiment in the tech industry.

she isn't just an innocent programmer, she writes about and is paid to give speeches about very clear political stances.

self censorship is not the way.

>>885419

This is a cuck mindset. Don't hide this attitude. OVERWHELM them with it.

Wow, other people have to learn what you already know, what a scandal!

>>885528

you have to go back safia

>>885918

No, this anon is right. Nobody really cares. All the more reason this is retarded shilling is completely asinine.

>>885427 see

>>886154

Also, your premise doesn't hold up. It's not that she is now on an equal footing to me; it's that now that she's equal to millions of us in a single, low-effort manner and this little fact somehow springboards her to fame across the internet. We all know what a syscall is but she somehow does what any of us did as shitty little teenagers and suddenly she's a fucking tech godess. You don't see the problem here?

>>881231

>Hipster language

Can you show me where on this doll the hipsters touched your programming language?

>>885258

>It was one of the first of many instances where men crossed the line between professional and romantic.

The grammar's correct, but that sentence just feels wrong.

>>881485

notice she started the interships a few months before she went to Uni.

>>881276

Had you heard about numpy, OpenCV, OpenCL for example?

>>881314

re. 3 --- well you're free to try to do math stuff in sh. tell us when you succeed.

re. 2 and 1 --- this is a pure bullshit

>>892509

That's what he said

>>885417

There aren't many that diversity hires actually. It's just that they're so loud (which is their purpose) that it seems they're somehow a much larger minority.

>>881248

>worked as intern while attending university

Is this not standard practice is burgerland? I got paid internships at uni from 2nd year here.

>>881203

11 thousands tweet is a perfectly normal number for anyone using social networks these days. 4-5 tweets a day is in fact NOTHING, considering each tweet is just a witty phrase you can send from your NSAphone.

Why are you obsessed with this girl?

>>892520

why are you bumping the previously dead thread?

you have to go back safia

stop bumping this shit

>Why is shit like this getting circulated?

Why are you circulating it?

>>900950

Fuck off shill

>>880425

That's just what britbongs look like.