Coding Horrors: You can gather all your technical debt into one easy framework!

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Cavern of COBOL > Coding Horrors: You can gather all your technical debt into one easy framework!

«‹›1502 »

Edison was a dick: Apr 3, 2010; direct current only

geonetix posted:

I'm not actually sure why, it's probably because I learned to computer that way. And I cannot live without "bindkey -v" to enable vi-like editing. If somebody has an actual valid pro/cons table of zsh vs bash vs (t)csh I'll be the first to read it.

code:

set -o vi

promises to do that in bash, but I've not done that in zsh so I can't compare.

Also I found csh considered harmful a while back, but never having used csh, I can't say how valid it is.

Edison was a dick fucked around with this message at 23:15 on Jul 24, 2012

# ? Jul 24, 2012 23:09

Adbot: ADBOT LOVES YOU

# ? Apr 29, 2024 16:20

theratking: Jan 18, 2012

love this poo poo.

# ? Jul 24, 2012 23:43

Sinestro: Oct 31, 2010; The perfect day needs the perfect set of wheels.

This is what happens if you let a front-end guy write your Django app.

Python code:

from django.db import models

class Data(models.Model):
    key = models.TextField()
    value = models.TextField()

You are looking at the entire models.py. :suicide:

# ? Jul 25, 2012 00:40

Doc Hawkins: Jun 15, 2010; Dashing? But I'm not even moving!

theratking posted:

love this poo poo.

Github issues are pretty bad, but I wouldn't award them full-blown horror status

# ? Jul 25, 2012 00:57

how!!: Nov 19, 2011; by angerbot

Sinestro posted:

This is what happens if you let a front-end guy write your Django app.
Python code:
from django.db import models

class Data(models.Model):
    key = models.TextField()
    value = models.TextField()
You are looking at the entire models.py.

I recently phone interviewed with a company who gave me access to their github, and their code was exactly like this. They had one model called DictionaryStruct which was exactly like the one you posted, which was used for everything. They offered me a very good salary, but I had to decline their offer.

# ? Jul 25, 2012 02:05

Zamujasa: Oct 27, 2010; Bread Liar

Sinestro posted:

This is what happens if you let a front-end guy write your Django app.
Python code:
from django.db import models

class Data(models.Model):
    key = models.TextField()
    value = models.TextField()
You are looking at the entire models.py.

Boss did something like this (I think) in C#, where every single "global" value he created was its own class with a single private property, with the exact same get/set code (that is, copy/return the value, nothing else).

I combined them all into one class called GlobalVars and accessed the properties from there.

Not really a horror to use global variables in this circumstance, since we were wrapping around another app, and the whole disaster ran for a few seconds to get some data and died.

Aside: Bosscode has taken the "pay less, pay often" method of coding again, adding code to set the default timezone to every file instead of just setting the drat setting in php.ini once.

Then again, this the same person who copy-pastes the same "header" (comprised of two function definitions that are the exact same, but with a number in the name of one of them), so... :rolleyes:

# ? Jul 25, 2012 03:49

theratking: Jan 18, 2012

Doc Hawkins posted:

Github issues are pretty bad, but I wouldn't award them full-blown horror status

I'd say the horror is that my direct supervisor doesn't really fully understand the concept of an error or an error message. His response to "what's the 500 error" is "it's a 500 error."

# ? Jul 25, 2012 07:34

ymgve: Jan 2, 2004; Offensive Clock

I don't even know what it's about, but some browsers try to be "smart" by replacing HTTP errors with their own "friendly" error messages. It's possible that 500 error is all that he sees and there are no specific details shown.

edit: And by "some browsers" I mean "At least Internet Explorer 9", which gives me this

on this HTTP error test page

ymgve fucked around with this message at 08:01 on Jul 25, 2012

# ? Jul 25, 2012 07:57

Impotence: Nov 8, 2010; Lipstick Apathy

# ? Jul 25, 2012 08:02

Jewel: May 2, 2009

Biowarfare posted:

Funnily enough, I think chrome had the same thing. I think it was a 404 page I was trying to get working with a .htaccess on an Apache server, but it wasn't working. Turns out it had to be over a certain filesize for it to work, and just placing a single <img> wasn't big enough :wtc:

# ? Jul 25, 2012 08:11

theratking: Jan 18, 2012

I thought he knew what he was doing because he was directly reading the response from chrome dev-tools.

# ? Jul 25, 2012 10:31

Deus Rex: Mar 5, 2005

Doc Hawkins posted:

Github issues are pretty bad, but I wouldn't award them full-blown horror status

why are github issues bad. they do what they're supposed to pretty well i think

# ? Jul 25, 2012 10:46

New Yorp New Yorp: Jul 18, 2003; Only in Kenya.; Pillbug

Here's a nice, simple horror.

From the source of a product thumbnail page, repeated for every single thumbnail present on the page (minimum 36):

code:

        <!-- PM - STAR RATINGS GO HERE 

            

            TODO: the width for bn_current_rating needs to be calculated as a percentage (i.e. Rating 3.5 * 20 (each star is worth 20))

            See the script calc_rating in global.js to see how the calculation works



            <div class="bn_star_container">

	            <ul class="bn_star_rating">

		            <li class="bn_current_rating" style="width:80%;"></li> 

		        </ul>

	        </div>

        -->

Horror #2 is that they don't minify their JavaScript.

# ? Jul 25, 2012 14:29

Pochoclo: Feb 4, 2008; No...; Clapping Larry

The most horrendous code I've ever seen, is some guy that thought he was coding motherfucking assembly code in PHP or something, check this out:

code:

$get_data = strpos($url_file,'-');
		if ($get_data!==false) {
			$url_src = $url_file;
			$url_file = substr($url_src,0,$get_data);	// Se actualiza el valor de $url_file
			$DX = substr($url_src,$get_data+1,strlen($url_src));
			$url_src = $DX;
			
			$get_data = strpos($url_src,'-');
			if ($get_data!==false) {
				
				$DX = substr($DX,0,$get_data);
				$EX = substr($url_src,$get_data+1,strlen($url_src));
				$url_src = $EX;
				$get_data = strpos($url_src,'-');
				if ($get_data!==false) {
					
					$EX = substr($EX,0,$get_data);
					$FX = substr($url_src,$get_data+1,strlen($url_src));
					$url_src = $FX;
					$get_data = strpos($url_src,'-');
					if ($get_data!==false) {
						
						$FX = substr($FX,0,$get_data);
						$GX = substr($url_src,$get_data+1,strlen($url_src));
						$url_src = $GX;
						$get_data = strpos($url_src,'-');
						if ($get_data!==false) {
							
							$GX = substr($GX,0,$get_data);
							$HX = substr($url_src,$get_data+1,strlen($url_src));
							$url_src = $HX;
							$HX = FriendlyUrlDecode($HX);
						}
						$GX = FriendlyUrlDecode($GX);
					}
					$FX = FriendlyUrlDecode($FX);
				}
				$EX = FriendlyUrlDecode($EX);
			}
			$DX = FriendlyUrlDecode($DX);
			$x_vars_set = true;
		}

Yeah, did you see those variable names?
The entire loving CMS is a single clusterfuck of an index.php, which has mySQL queries, HTML, and basically everything in it. It's also 3000 lines long.
ALL VARIABLE NAMES ARE TWO LETTERS LONG.

It's madness.

I mean, here I am, making a real effort to keep my team following best practices, coding in symfony2 neatly (you may or may not agree with the choice of symfony2, but it's still a very neat and modern MVC framework, I like it very much)... and there's still guys like THAT who are seemingly stuck in 1999 or something. What the flying gently caress.

On the other hand, thanks to them it's really easy to get a good-paying job. Companies here are so disappointed with the rampant bad coders, they'll hire a semi-competent programmer faster than you can blink.

# ? Jul 25, 2012 16:54

Impotence: Nov 8, 2010; Lipstick Apathy

Ithaqua posted:

Here's a nice, simple horror.

From the source of a product thumbnail page, repeated for every single thumbnail present on the page (minimum 36):

code:

        <!-- PM - STAR RATINGS GO HERE 

            

            TODO: the width for bn_current_rating needs to be calculated as a percentage (i.e. Rating 3.5 * 20 (each star is worth 20))

            See the script calc_rating in global.js to see how the calculation works



            <div class="bn_star_container">

	            <ul class="bn_star_rating">

		            <li class="bn_current_rating" style="width:80%;"></li> 

		        </ul>

	        </div>

        -->

Horror #2 is that they don't minify their JavaScript.

Webserver I use automatically strips non-[if IE] comments and minifies CSS and JS automatically :v:

# ? Jul 25, 2012 17:29

Beef: Jul 26, 2004

Pochoclo posted:

The most horrendous code I've ever seen, is some guy that thought he was coding motherfucking assembly code in PHP or something, check this out:

I find the repeating nested program structure to be a much worse coding horror than the 2-letter variables.

# ? Jul 25, 2012 17:46

Pochoclo: Feb 4, 2008; No...; Clapping Larry

Beef posted:

I find the repeating nested program structure to be a much worse coding horror than the 2-letter variables.

Yeah, there's a reason I picked that specific part.

Now imagine all the 3000 lines of it are just as terrible.

It's like Boatmurdered, in PHP form.

# ? Jul 25, 2012 19:08

Golbez: Oct 9, 2002; 1 2 3!
If you want to take a shot at me get in line, line
1 2 3!
Baby, I've had all my shots and I'm fine

PHP, is_readable(), has a curiosity in its documentation.

quote:

bool is_readable ( string $filename )

Tells whether a file exists and is readable.

Returns TRUE if the file or directory specified by filename exists and is readable, FALSE otherwise.

OK? So you run it on a filename and it tells you if it both exists, and is in a readable permission. Awesome.

Except...

quote:

Upon failure, an E_WARNING is emitted.

I just tested it on a nonexistent file and it return false and did not throw a warning. So what on earth does 'upon failure' mean? What situation could a built-in function like this fail that doesn't count as a fatal error?

# ? Jul 25, 2012 20:35

leterip: Aug 25, 2004

Golbez posted:

PHP, is_readable(), has a curiosity in its documentation.

OK? So you run it on a filename and it tells you if it both exists, and is in a readable permission. Awesome.

Except...

I just tested it on a nonexistent file and it return false and did not throw a warning. So what on earth does 'upon failure' mean? What situation could a built-in function like this fail that doesn't count as a fatal error?

The bigger horror is that it exists at all. There is no way to provide that function without creating a race condition. What if someone deletes the file in between that call and your attempt to read it?

# ? Jul 25, 2012 20:48

Golbez: Oct 9, 2002; 1 2 3!
If you want to take a shot at me get in line, line
1 2 3!
Baby, I've had all my shots and I'm fine

leterip posted:

The bigger horror is that it exists at all. There is no way to provide that function without creating a race condition. What if someone deletes the file in between that call and your attempt to read it?

And if I try to run fopen() on a non-existent file, it throws a warning. Keep in mind that it also returns false. The documentation even helpfully suggests using @ to suppress the warning. Perhaps PHP functions should throw exceptions, because this seems like a reasonable place for that.

# ? Jul 25, 2012 21:03

pseudorandom name: May 6, 2007

Golbez posted:

PHP, is_readable(), has a curiosity in its documentation.

OK? So you run it on a filename and it tells you if it both exists, and is in a readable permission. Awesome.

Except...

I just tested it on a nonexistent file and it return false and did not throw a warning. So what on earth does 'upon failure' mean? What situation could a built-in function like this fail that doesn't count as a fatal error?

In theory, all the usual problems you'd have with a file name (too much symbolic link indirection, path name is too long, one of the directories in the path isn't, etc.), except PHP converts any error returned from access() to false and calls it a day.

# ? Jul 25, 2012 21:08

leterip: Aug 25, 2004

Golbez posted:

And if I try to run fopen() on a non-existent file, it throws a warning. Keep in mind that it also returns false. The documentation even helpfully suggests using @ to suppress the warning. Perhaps PHP functions should throw exceptions, because this seems like a reasonable place for that.

The point is that you have to assume that the fopen() call could raise a warning regardless of how much checking you do because by the time you get to actually opening the file the state of the filesystem could be different, so why bother doing the checking? It's a completely worthless function.

# ? Jul 25, 2012 21:38

raminasi: Jan 25, 2005; a last drink with no ice

leterip posted:

The bigger horror is that it exists at all. There is no way to provide that function without creating a race condition. What if someone deletes the file in between that call and your attempt to read it?

That isn't exactly a PHP problem, though. Tons of APIs have stuff like that.

# ? Jul 25, 2012 21:52

Sebbe: Feb 29, 2004

I've been lobbying to the PHP developers for years to change is_readable() to take source code as input instead of a file name. :v:

# ? Jul 25, 2012 23:01

Gazpacho: Jun 18, 2004; by Fluffdaddy; Slippery Tilde

Gazpacho posted:

I can assure you that the real problem is programmers just not knowing what memory is.

Followup: http://blogs.msdn.com/b/oldnewthing/archive/2012/07/25/10333134.aspx

# ? Jul 26, 2012 02:48

Strong Sauce: Jul 2, 2003; You know I am not really your father.

Maybe doesn't belong in here, but...

Only registered members can see post attachments!

# ? Jul 26, 2012 04:52

hobbesmaster: Jan 28, 2008

Strong Sauce posted:

Maybe doesn't belong in here, but...

I assume that the second line 27 is supposed to be a *?

# ? Jul 26, 2012 05:47

pokeyman: Nov 26, 2006; That elephant ate my entire platoon.

TextMate 2 is if not a coding horror at least a coding project horror. Is it possible to parlay years of widespread success into something more flaccid than TM2?

# ? Jul 26, 2012 07:11

Strong Sauce: Jul 2, 2003; You know I am not really your father.

I'm still using TextMate and those were two separate lines, not a wordwrapped line. Not sure what caused the problem.

# ? Jul 26, 2012 07:51

Comrade Gritty: Sep 19, 2011; This Machine Kills Fascists

leterip posted:

The point is that you have to assume that the fopen() call could raise a warning regardless of how much checking you do because by the time you get to actually opening the file the state of the filesystem could be different, so why bother doing the checking? It's a completely worthless function.

There are other reasons to check if a file exists or is readable than to write (or read) from it. Sometimes you just want to know if the file exists and is readable.

# ? Jul 27, 2012 11:10

ozymandOS: Jun 9, 2004

Steampunk Hitler posted:

There are other reasons to check if a file exists or is readable than to write (or read) from it. Sometimes you just want to know if the file exists and is readable.

Okay. But the information you receive from this function could be immediately out of date, no matter what you are going to use it for. It's a useless function.

# ? Jul 27, 2012 14:27

silvergoose: Mar 18, 2006; IT IS SAID THE TEARS OF THE BWEENIX CAN HEAL ALL WOUNDS

BP posted:

Okay. But the information you receive from this function could be immediately out of date, no matter what you are going to use it for. It's a useless function.

If all you want to know is if a process ran already today, you can just check to see if a file is there and readable. If someone for some godawful reason deletes it thereafter to try to have the process run again anyway, who the hell cares, you don't run again and it's their fault for being slow.

(this is for the very popular "touch ALREADY_RAN_DATESTAMP" way of tracking this, of course)

# ? Jul 27, 2012 14:33

Golbez: Oct 9, 2002; 1 2 3!
If you want to take a shot at me get in line, line
1 2 3!
Baby, I've had all my shots and I'm fine

^^^ okay I suppose that works

BP posted:

Okay. But the information you receive from this function could be immediately out of date, no matter what you are going to use it for. It's a useless function.

What about

Scenario 1: Have a script that runs for 10 minutes but only if it can read a file. So we check if the file is readable first, then run the 10 minutes, and at the end open the file and write to it.

Scenario 2: Open the file first because handles are cheap and it doesn't matter if you fopen it or check if it's readable, but this keeps it locked for 10 minutes which really is that a bad thing?

Yeah, you're right, I got nothing.

So, then we have is_writable(), but also this nugget: file_exists().

quote:

bool file_exists ( string $filename )

Checks whether a file or directory exists.

Returns TRUE if the file or directory specified by filename exists; FALSE otherwise.

Note: The check is done using the real UID/GID instead of the effective one.

Note: Because PHP's integer type is signed and many platforms use 32bit integers, some filesystem functions may return unexpected results for files which are larger than 2GB.

I am probably missing something here, but why the hell would this return "unexpected results" because of a large file? It can't see if that file exists if it's more than 2gb? What unexpected results would we have? Returning false if the file is actually there? Is this actually a thing that happens?

Golbez fucked around with this message at 14:58 on Jul 27, 2012

# ? Jul 27, 2012 14:36

Plorkyeran: Mar 22, 2007; To Escape The Shackles Of The Old Forums, We Must Reject The Tribal Negativity He Endorsed

Golbez posted:

I am probably missing something here, but why the hell would this return "unexpected results" because of a large file? It can't see if that file exists if it's more than 2gb? What unexpected results would we have? Returning false if the file is actually there? Is this actually a thing that happens?

I assume it's just a generic warning that a monkey slapped on all filesystem functions rather than applying any sort of thought.

Then again it is PHP, so maybe checking for the existence of a file includes checking the size of the file for some reason.

# ? Jul 27, 2012 15:04

Goat Bastard: Oct 20, 2004

BP posted:

Okay. But the information you receive from this function could be immediately out of date, no matter what you are going to use it for. It's a useless function.

Well "this file existed when I started doing what I'm doing so I'm going to assume it still exists" and "this file has never existed within the scope of what I'm doing" are two quite different situations that usually call for different actions. Obviously you still have to write defensive code, but it's a bit of stretch to say that checking if a file exists is useless.

# ? Jul 27, 2012 16:06

ozymandOS: Jun 9, 2004

Goat Bastard posted:

Well "this file existed when I started doing what I'm doing so I'm going to assume it still exists" and "this file has never existed within the scope of what I'm doing" are two quite different situations that usually call for different actions. Obviously you still have to write defensive code, but it's a bit of stretch to say that checking if a file exists is useless.

True. I guess you can use a file existence check as a way to short-circuit unnecessary code--ie skip doing processing that will eventually fail due to a missing file, rather than waiting until you actually need the file before failing. But you should always also check at time of open.

# ? Jul 27, 2012 17:23

Zamujasa: Oct 27, 2010; Bread Liar

silvergoose posted:

If all you want to know is if a process ran already today, you can just check to see if a file is there and readable. If someone for some godawful reason deletes it thereafter to try to have the process run again anyway, who the hell cares, you don't run again and it's their fault for being slow.

(this is for the very popular "touch ALREADY_RAN_DATESTAMP" way of tracking this, of course)

I use something like this for a cron job at work; it only runs every 5 minutes (or manually), and when it starts up it checks if a file exists (and tries to get the content if it does). If the file doesn't exist, it continues. If the file does exist, but the timestamp is more than five minutes old, it continues as well (but writes the latest timestamp to the file). At the end, the file is removed.

It's not entirely worthless; I personally think that the obsession over race conditions is a little unwarranted if you have some control over what's going on and you aren't multithreaded, but everybody seems to throw a fit over them.

# ? Jul 27, 2012 17:34

Eggnogium: Jun 1, 2010; Never give an inch! Hnnnghhhhhh!

Zamujasa posted:

I use something like this for a cron job at work; it only runs every 5 minutes (or manually), and when it starts up it checks if a file exists (and tries to get the content if it does). If the file doesn't exist, it continues. If the file does exist, but the timestamp is more than five minutes old, it continues as well (but writes the latest timestamp to the file). At the end, the file is removed.

It's not entirely worthless; I personally think that the obsession over race conditions is a little unwarranted if you have some control over what's going on and you aren't multithreaded, but everybody seems to throw a fit over them.

A little caution is warranted even when today's environment doesn't call for it, because next year all kinds of other processes could have been introduced that interfere with your program's resources.

# ? Jul 27, 2012 18:15

Golbez: Oct 9, 2002; 1 2 3!
If you want to take a shot at me get in line, line
1 2 3!
Baby, I've had all my shots and I'm fine

We're losing sight of the big picture. The existence of is_readable is not the true horror. The true horror is that fopen() will always throw a warning if the file is not readable, unless you silence errors on the function call. It seems like they expect you to run is_readable first, and then run fopen, but in normal usage that could still result in fopen not opening the file. Most people would check to see if it returned false and move on, but now you have to manipulate the error reporting or simply ignore when your log gets a warning, both of which are poor options.

# ? Jul 27, 2012 18:17

Adbot: ADBOT LOVES YOU

# ? Apr 29, 2024 16:20

Zamujasa: Oct 27, 2010; Bread Liar

Eggnogium posted:

A little caution is warranted even when today's environment doesn't call for it, because next year all kinds of other processes could have been introduced that interfere with your program's resources.

I agree, I'm just saying that everyone seems to be acting as if race conditions are the newest goto.

As for PHP, whatever happened to PHP 6? I seem to recall it existing at some point years and years ago, but then it got rolled back out of nowhere and seems to have been erased from existence. Is that just my imagination?

Because it seems like a PHP 6 would be a great time to start working on cleaning up some of the larger horrors in PHP, like the lack of exceptions.

# ? Jul 27, 2012 18:20

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Cavern of COBOL > Coding Horrors: You can gather all your technical debt into one easy framework!

«‹›1502 »