The PL (Programming Language) thread: I'm glad Python is the current hipste

«‹›14 »

The Something Awful Forums > Discussion > Serious Hardware / Software Crap > YOSPOS > The PL (Programming Language) thread: I'm glad Python is the current hipste

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

never feed perl5 after midnight

# ¿ Apr 28, 2012 00:34

Adbot: ADBOT LOVES YOU

# ¿ May 23, 2013 05:35

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Gentle Autist posted:

Programmers > 40 how much has your job changed over the course of your career

Did you ever consciously have to go learn something off the clock to make a career move or did you just change jobs/roles as they came up

i bet my career on desktop development, then over the last 10 years it has all gone away leaving me literally in the same career position as a trained photofinisher, beating my head against recruiting agencies who want to know if i can agile in agile

# ¿ Apr 28, 2012 03:43

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

z0ratio fartboner posted:

owned

yep

learn from my story

# ¿ Apr 28, 2012 03:57

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Inverse Icarus posted:

which particular things are you complaining about? like what browser plugin do you think should be a standalone app but isn't?

it's not a matter of network vs. standalone, network apps were going to happen anyway, it's about people building inner platforms and thinking they're awesome for doing so

# ¿ Apr 28, 2012 04:31

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Tiny Bug Child posted:

like how

# ¿ Apr 28, 2012 04:37

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

HTML5 posted:

or it's because people don't do just one thing at a time anymore

lol the overwhelming majority of windows users didn't know how to switch windows and now they don't know or care what tabs are, multitasking is and always will be a nerd feature

# ¿ Apr 28, 2012 19:37

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

ps sorry bout ruining your languages thread op

# ¿ Apr 28, 2012 19:39

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

here let me try to get back to the subject

# ¿ Apr 28, 2012 20:01

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

if "engineers" had their way every computer application would literally be a labview project

# ¿ Apr 28, 2012 21:03

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

womynizer posted:

i see your point, but masking the features most common for people that make the most out of your application? it's like hiding the steering wheel in a racing game

whoa what applications are you talking about

# ¿ Apr 28, 2012 21:16

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

btw i don't know what racing games you've been playing but most of them are not controlled by dragging a steering wheel around

# ¿ Apr 28, 2012 21:18

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

pyf monad for automatically inserting nextTick calls in a CPU-bound calculation

# ¿ Apr 29, 2012 07:12

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

it is perhaps short-sighted to blame windows specifically because windows was a product of the microcomputer hobbyist culture which itself rejected large-scale computing, the real problem is systems becoming so complex and legacy-bound that they are no longer hackable

this of course implies that eventually we'll see a movement to create inner platforms within the web/mobile platforms

Gazpacho fucked around with this message at Apr 29, 2012 around 07:27

# ¿ Apr 29, 2012 07:20

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

tef posted:

but threads are hard

cooperative timeslicing isn't any more fun, or is there anyone left who still remembers developing win16?

Gazpacho fucked around with this message at Apr 29, 2012 around 10:52

# ¿ Apr 29, 2012 10:50

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

TiMBuS posted:

unless the thing you are waiting for seriously refuses to not block. in which case i guess they suck

it sucks for basically any long-running algorithm, you end up adding yield calls and/or throwing yourself continuation events where it seems appropriate which is dual to adding locks in a preemptive system. web workers solve this but if they have been adopted in nodejs i havent heard about it

Gazpacho fucked around with this message at Apr 29, 2012 around 11:15

# ¿ Apr 29, 2012 11:10

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

rotor posted:

lmao if your program isn't simply a collection of smaller single-purpose programs that talk to each other using character streams that can have their parallelism managed by the os

i mostly agree but serializing data to a byte stream for IPC is a dumb, shared memory passing ftw

# ¿ Apr 30, 2012 04:21

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Janin posted:

this is easily solved by spending more than like twenty minutes learning the language

worse is better, you will never sell a language that depends on a graduate CS education to get started vs. one in which you can write "og say print stuff to screen"

# ¿ Apr 30, 2012 05:24

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Janin posted:

code:

main = do
    putStrLn "uggh how og print"
    putStrLn "og confused without semicolons, og smash"

sweet now try doing it inside an arbitrary function OH SNAP

# ¿ Apr 30, 2012 06:02

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

UML is yesterday's buzzword, agile my agile pls

# ¿ Apr 30, 2012 09:14

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Gazpacho posted:

# ¿ Apr 30, 2012 10:04

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Janin posted:

why would you want to do this? IO should only be done if the function's type signature allows IO.

lol, "should" in one hand and write imperative code in the other and tell me which fills up the fastest

# ¿ Apr 30, 2012 18:55

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

newreply.php posted:

lets just talk about languages that make the most money for ppl knowing them

p. sure that's vb or maybe cobol

# ¿ Apr 30, 2012 19:28

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Rufo posted:

maybe u should have gone to a school for people instead

# ¿ Apr 30, 2012 23:12

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

HORATIO HORNBLOWER posted:

except for the total bullshit 400 level mccarthy witch hunt english seminars

wait what

# ¿ May 1, 2012 03:49

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

ahhh spiders posted:

it actually knows what the problem is, though

error: `>>' should be `> >' within a nested template argument list

apparently the c++ committee had long discussions about that before someone finally mentioned that instead of throwing an error they could tweak the parser state to accept it

# ¿ May 1, 2012 19:58

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Werthog posted:

having trouble understanding why this would possibly be the correct behavior unless there was a valid way of using the stream operator within a nested template argument list, apart from spergy "when in doubt, fail" attitudes

dogmatic distinction between lexing and parsing

# ¿ May 1, 2012 20:01

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

JawnV6 posted:

youre probably right, the recovery code for that must look terrible

# ¿ May 1, 2012 20:17

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

rotor posted:

has anyone ever written forth? it seems so neat.

it's a neat idea but i've met some of chuck moore's clique and they're pretty cultish, always talking about the great things that could be done with forth except that the "forth philosophy" says you shouldn't

i suppose if you wants to write actual programs in 100% continuation passing style by hand that's the language to do it in

# ¿ May 2, 2012 04:46

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

btw i'm not really talking about ansi forth, i'm talking about any one of chuck moore's custom forths like this one that replaces the OS and color highlighting is operationally significant and it has its own optimized keyboard layout

also here's a photo of chuck which would make a good avatar

# ¿ May 2, 2012 04:58

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Gazpacho posted:

i suppose if you wants to write actual programs in 100% continuation passing style by hand that's the language to do it in

I thought i'd put in a little

and write up what I mean. I've had this in the back of my mind for a few weeks but I've never seen it discussed online because Forth is a kind of ghetto language to academics. Take your naive Fibonacci in Javascript.

code:

function fib(n) {
  if(n < 2)
    return 1;
  else {
    return fib(n - 1) + fib(n - 2);
}

Next transform just the fib function to continuation style. Note that it's starting to look like postfix because you have the two recursive fibs first then the addition.

code:

function fib(n, cont) {
  if(n < 2)
    cont(1);
  else {
    fib(n - 1, function(x) {
      fib(n - 2, function(y) {
        cont(x + y);
      });
    });
  }
}

Now go all out and transform the operators and control statements as well.

code:

function fib(n, cont) {
  op<(n, 2, function(x0) {
    ifelse(x0,
      function() { cont(1); },
      function() {
        op-(n, 1, function(x1) {
          fib(x1, function(x2) {
            op-(n, 2, function(x3) {
              fib(x3, function(x4) {
                op+(x2, x4, cont);
              });
            });
          });
        });
      });
    });
}

From this you can read off the corresponding Forth code line by line.

code:

// ": fib" (define fib)
function fib(n, cont) {
  // "dup" (preserve n for later use) "2 <"
  op<(n, 2, function(x0) {
    // "if"
    ifelse(x0,
      // "drop" (n not used on this branch) "1"
      function() { cont(1); },
      // "else"
      function() {
        // "dup" (preserve n) "1 -"
        op-(n, 1, function(x1) {
          // "fib"
          fib(x1, function(x2) {
            // "swap" (reach up the stack past x2 to get n for last use) "2 -"
            op-(n, 2, function(x3) {
              // "fib"
              fib(x3, function(x4) {
                // "+ then" (end of false branch) ";" (call cont)
                op+(x2, x4, cont);
...

Result:

code:

: fib dup 2 < if drop 1 else dup 1 - fib swap 2 - fib + then ;

There's a quirk here that the second cont call translates to ; but the first doesn't. colorForth actually handles this consistently so "else" becomes ";".

Gazpacho fucked around with this message at May 2, 2012 around 10:45

# ¿ May 2, 2012 10:40

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

tef posted:

node.js

haha no i mean ALL THE FUKKEN WAY

e: dammit why did i post that at the end of the page

# ¿ May 2, 2012 11:14

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

well i guess we've settled it then, forth is the cool programming language for stuff

# ¿ May 2, 2012 19:39

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

at least javascript pushed closures into the mainstream, that only took what 45 years?

# ¿ May 3, 2012 02:32

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

remember when you could script a ticker message in the browser status bar? i miss that

# ¿ May 3, 2012 02:37

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Gentle Autist posted:

lol if you don't have shaggar on ignore

hes right about the web platform tho

maybe if microsoft hadn't ridden c++ into the ground as an app language before creating .net a few years too late...

# ¿ May 3, 2012 03:41

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

a month ago i got a recruiter contact about a web dev position and the recruiter sent me an employer questionnaire with a question that was more or less "why is nodejs generally recognized as the coolest new technology around?" i had never heard of nodejs before so that was my first run-in with its fans

# ¿ May 3, 2012 04:28

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

Inverse Icarus posted:

this is me

hello

tbf there's lots of EEs who understand circuits and digital logic perfectly well but they glaze over at standard CS theoretical concepts

its like that line in hitchhiker's guide, "they still think digital watches are pretty neat"

(this includes university professors btw)

# ¿ May 4, 2012 01:46

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

when you get into category theory you find out that calculus concepts have their equivalents in the discrete realm that have applications but really who ever does that

e.g. you can differentiate or integrate a data type and the result is useful

# ¿ May 5, 2012 01:45

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

PENETRATION TESTS posted:

i wrote a CGI app in R, it literally ran a new R process for every request and is the most horrifying thing ever

CGI is a perfectly fine way to run code

# ¿ May 7, 2012 04:15

Adbot: ADBOT LOVES YOU

# ¿ May 23, 2013 05:35

Gazpacho: Jun 18, 2004; 2 MEGA 2 FAIL

using stringbuffer/builder when it makes sense to use it is like the second thing people are taught in java so i dont know what you're spergin about

# ¿ May 7, 2012 23:03

«‹›14 »

The Something Awful Forums > Discussion > Serious Hardware / Software Crap > YOSPOS > The PL (Programming Language) thread: I'm glad Python is the current hipste