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!

«‹›1501 »

1337JiveTurkey: Feb 17, 2005

Bognar posted:

If we're gonna shake up the front-end language options, can we at least pick a language with type checking?

Since nobody's going to use it directly anyhow, I'd just go with a more secure subset of the JVM with none of the dangerous bits to begin with.

# ? Jun 27, 2016 18:08

Adbot: ADBOT LOVES YOU

# ? Apr 23, 2024 11:56

fritz: Jul 26, 2003

TimWinter posted:

Why can't browsers just run python.

I think there's a question of what comes first, bad languages or bad programmers, and in the case of web stuff I have never seen any indication that it's not the second option.

# ? Jun 27, 2016 18:16

Pollyanna: Mar 5, 2005; Milk's on them.

Update on the help button: we decided to place a link to the user guide under the pop-up search bar for finding people. :pwn:

# ? Jun 27, 2016 18:20

nielsm: Jun 1, 2009

Pollyanna posted:

Update on the help button: we decided to place a link to the user guide under the pop-up search bar for finding people.

Sounds like the equivalent of building the bike shed underground, with a trapdoor hidden in the woods as the only access.

# ? Jun 27, 2016 18:33

HappyHippo: Nov 19, 2003; Do you have an Air Miles Card?

Pollyanna posted:

Also we're having a 30 minute meeting on where to put the "Get Help" button.

Bikeshedding.txt

# ? Jun 27, 2016 18:34

JawnV6: Jul 4, 2004; So hot ...

C's a lot of fun, but I don't really think of it as squeezing performance. More like making peek/poke machine accesses prettier.

Anyway, gcc 7.0.0 optimizations are breaking the Kernel build.

nielsm posted:

Sounds like the equivalent of building the bike shed underground, with a trapdoor hidden in the woods as the only access.

Which will now become the user's expectation and must not ever be moved.

# ? Jun 27, 2016 18:44

Space Kablooey: May 6, 2009

Pollyanna posted:

Update on the help button: we decided to place a link to the user guide under the pop-up search bar for finding people.

At that point, why even bother?

# ? Jun 27, 2016 18:55

Soricidus: Oct 21, 2010; freedom-hating statist shill

Nobody's ever going to read the user guide anyway

# ? Jun 27, 2016 19:03

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

Hammerite posted:

You may think it's reasonable to have objects in the global namespace have effects on script execution based on magic names that aren't visually cued in any way, and call this opinion "experience", but I don't think it's reasonable, and I call it "brain damage caused by too much Javascript exposure"

yeah javascript is trash in that way. maybe next time spend some effort to read up so you don't run into these problems? but it's cool, you'll know better next time.

# ? Jun 27, 2016 19:23

canis minor: May 4, 2011

My favorite JS weirdness is naming a form submit button "submit" and then trying to invoke submit call on such form - as form elements are referenced as [form].[name of element] naming an element this way overwrites submit event on the form. Works on other properties as well!

# ? Jun 27, 2016 20:02

Xarn: Jun 26, 2015

My favourite language is actually C++. :v:

# ? Jun 27, 2016 21:10

Dirty Frank: Jul 8, 2004

Xarn posted:

My favourite language is actually C++.

Are you looking for a dom? ASL?

# ? Jun 27, 2016 21:23

leper khan: Dec 28, 2010; Honest to god thinks Half Life 2 is a bad game. But at least he likes Monster Hunter.

Dirty Frank posted:

Are you looking for a dom? ASL?

I thought people looking for the dom wrote JavaScript. :v:

# ? Jun 27, 2016 21:53

The MUMPSorceress: Jan 6, 2012; ^SHTPSTS; Gary’s Answer

Dirty Frank posted:

Are you looking for a dom? ASL?

i write vb so I'm all about a private sub

# ? Jun 27, 2016 22:16

Internet Janitor: May 17, 2008; "That isn't the appropriate trash receptacle."

TooMuchAbstraction posted:

We hear so much about how all these languages suck...so what are y'all's favorite programming languages? And perhaps more importantly, are your favorite languages the ones you actually get to use for any substantial serious work? No "I did a few Project Euler problems in Haskell and really liked it".

I really dig K. As of last week I'm getting paid to write it, too. Only time will tell whether I've earned myself my dream job or created my own personal hell. At least the pay's good and my coworkers all seem to be friendly, intelligent people. :unsmith:

# ? Jun 27, 2016 22:24

Amberskin: Dec 22, 2013; We come in peace! Legit!

TooMuchAbstraction posted:

We hear so much about how all these languages suck...so what are y'all's favorite programming languages? And perhaps more importantly, are your favorite languages the ones you actually get to use for any substantial serious work? No "I did a few Project Euler problems in Haskell and really liked it".

PL/1. Unironically.

# ? Jun 27, 2016 22:45

Hammerite: Mar 9, 2007; And you don't remember what I said here, either, but it was pompous and stupid.; Jade Ear Joe

Strong Sauce posted:

yeah javascript is trash in that way. maybe next time spend some effort to read up so you don't run into these problems? but it's cool, you'll know better next time.

Nah I'll just whine about it again, more cathartic and I can argue with people in this thread.

Ah, uh, I mean yes, I will study hard

# ? Jun 27, 2016 23:09

Malcolm XML: Aug 8, 2009; I always knew it would end like ｔｈｉｓ．

JawnV6 posted:

C's a lot of fun, but I don't really think of it as squeezing performance. More like making peek/poke machine accesses prettier.

Anyway, gcc 7.0.0 optimizations are breaking the Kernel build.

Which will now become the user's expectation and must not ever be moved.

More like the kernel relies on undefined behavior because C is underspecified

The real horror is that this kind of code can appear anywhere in big c codes libs

# ? Jun 28, 2016 01:25

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

Hammerite posted:

Nah I'll just whine about it again, more cathartic and I can argue with people in this thread.

Ah, uh, I mean yes, I will study hard

cool just don't get so butthurt when you post about your own code in a coding horrors thread and it gets ripped on.

# ? Jun 28, 2016 03:35

rjmccall: Sep 7, 2007; no worries friend; Fun Shoe

Internet Janitor posted:

The preprocessor, undefined behavior, typedef syntax (and for that matter having both concepts of typedefs and structs rather than unifying them into a single idea), the anemic build systems that inevitably result in people defining compile-time logic using ad-hoc bash scripts,

A lot of the undefined behavior in C/C++ is basically unavoidable without either a completely different execution model (which would make C useless for many of its intended use cases) or a wildly more advanced compiler (which would have been impossible for most of its lifetime) or in many cases both plus massive restrictions on the usability of the language.

I will not defend signed overflow being U.B., though. We should really just define that to be wraparound these days.

# ? Jun 28, 2016 03:48

Zemyla: Aug 6, 2008; I'll take her off your hands. Pleasure doing business with you!

rjmccall posted:

I will not defend signed overflow being U.B., though. We should really just define that to be wraparound these days.

What about the ones-complement and sign-magnitude integer machines that totally exist you guys so we need to keep them in the standard, just like trigraphs, because we make tons of machines that don't have square or curly brackets!

# ? Jun 28, 2016 05:27

Athas: Aug 6, 2007; fuck that joker

Zemyla posted:

What about the ones-complement and sign-magnitude integer machines that totally exist you guys so we need to keep them in the standard, just like trigraphs, because we make tons of machines that don't have square or curly brackets!

You can just make it implementation-defined instead. I suspect signed overflow is undefined behaviour because it allows the compiler to assume x+1>x if x is signed, which sometimes simplifies range analysis.

# ? Jun 28, 2016 06:49

Xarn: Jun 26, 2015

While I don't know of any sign-magnitude or ones-complement machine, aren't there still machines with saturating arithmetic? So defining it as overflowing in two-complement representation could still lead to pessimization on some hardware.

# ? Jun 28, 2016 09:53

feedmegin: Jul 30, 2008

TooMuchAbstraction posted:

We hear so much about how all these languages suck...so what are y'all's favorite programming languages?

The one I invented myself, of course.

Edit: on Windows, of course, it's not main() but WinMain(), because Microsoft gotta be different! :sun:

# ? Jun 28, 2016 11:22

nielsm: Jun 1, 2009

feedmegin posted:

The one I invented myself, of course.

Edit: on Windows, of course, it's not main() but WinMain(), because Microsoft gotta be different!

I don't know if any C compilers actually supported that directly, but you can definitely write a program that does one thing when run in a real DOS environment, and another thing when started from inside Windows <= 3.11. So I could imagine the compiler emitting an entry point that calls main() when run from DOS, and WinMain() when run from Windows.

# ? Jun 28, 2016 11:58

HappyHippo: Nov 19, 2003; Do you have an Air Miles Card?

Athas posted:

I suspect signed overflow is undefined behaviour because it allows the compiler to assume x+1>x if x is signed, which sometimes simplifies range analysis.

Yeah and this is what I mean by the standards being written for compiler writers as opposed to language users.

Also, my favorite example of undefined behaviour in c is

quote:

An unmatched � or � character is encountered on a logical source line during tokenization.

# ? Jun 28, 2016 13:17

Pollyanna: Mar 5, 2005; Milk's on them.

TooMuchAbstraction posted:

We hear so much about how all these languages suck...so what are y'all's favorite programming languages? And perhaps more importantly, are your favorite languages the ones you actually get to use for any substantial serious work? No "I did a few Project Euler problems in Haskell and really liked it".

Lisps! Specifically Clojure. It's fascinating to me and I have yet to crack the code on Lisp dialects' secrets to productivity and the "CODE IS DATA IS CODE" nirvana bullshit, and it may elude me for my entire life. But it sure sounds like a silver bullet.

Aside from that, I like Ruby and Elixir - I'm most productive in Ruby because that's where the grand majority of my experience lies, and I think about things in a functional manner most of the time and both Ruby and Elixir make that easy. As does Clojure, of course.

Not too hot on JavaScript, though ES6 is intriguing. .NET languages like C#, I don't get much chance to work in, which is a shame cause they don't look too bad.

Internet Janitor posted:

I really dig K. As of last week I'm getting paid to write it, too. Only time will tell whether I've earned myself my dream job or created my own personal hell. At least the pay's good and my coworkers all seem to be friendly, intelligent people.

Array languages like APL and K are also fascinating but I have no idea what I'd use them for :shrug:

Hard number crunching and data processing, right?

# ? Jun 28, 2016 14:57

feedmegin: Jul 30, 2008

nielsm posted:

I don't know if any C compilers actually supported that directly, but you can definitely write a program that does one thing when run in a real DOS environment, and another thing when started from inside Windows <= 3.11. So I could imagine the compiler emitting an entry point that calls main() when run from DOS, and WinMain() when run from Windows.

Nah. You're thinking of the MS-DOS stub at the start of PE/COFF files, which is completely separate and independent from the actual Windows program (and, of course, is running in 16-bit mode). You can't write one program which directly targets both environments or anything.

# ? Jun 28, 2016 15:03

nielsm: Jun 1, 2009

feedmegin posted:

Nah. You're thinking of the MS-DOS stub at the start of PE/COFF files, which is completely separate and independent from the actual Windows program (and, of course, is running in 16-bit mode). You can't write one program which directly targets both environments or anything.

I thought you could do more with that, i.e. make a file that works as both MZ and NE format. Guess not.

# ? Jun 28, 2016 15:19

fishmech: Jul 16, 2006; by VideoGames; Salad Prong

feedmegin posted:

Nah. You're thinking of the MS-DOS stub at the start of PE/COFF files, which is completely separate and independent from the actual Windows program (and, of course, is running in 16-bit mode). You can't write one program which directly targets both environments or anything.

You might not be able to write them in one source file, but there have been several programs out there that are both a full DOS program and a full 16 bit Windows program (or rarely 32 bit Windows program). It's usually only done to show off though.

# ? Jun 28, 2016 15:32

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

In a sense all windows programs were dual-mode programs, as the "you must run this in windows" message was printed by the program itself and not the loader.

# ? Jun 28, 2016 16:03

feedmegin: Jul 30, 2008

Plorkyeran posted:

In a sense all windows programs were dual-mode programs, as the "you must run this in windows" message was printed by the program itself and not the loader.

Yes, that would be the MS-DOS stub I just talked about. It's still there on modern Windows, even though for 64-bit Windows DOS isn't :shobon:

That said, I wasn't actually a Windows programmer back in the 3.1 days, so I'm willing to believe it was possible then in the pre-PE days.

feedmegin fucked around with this message at 16:21 on Jun 28, 2016

# ? Jun 28, 2016 16:19

Zemyla: Aug 6, 2008; I'll take her off your hands. Pleasure doing business with you!

I know regedit and Scandisk were among the programs that worked in both DOS and Windows.

# ? Jun 28, 2016 16:33

fishmech: Jul 16, 2006; by VideoGames; Salad Prong

Zemyla posted:

I know regedit and Scandisk were among the programs that worked in both DOS and Windows.

In Windows 95/98, Scandisk in Windows is accessed through a 5 kilobyte stub exe (scandskw.exe). Not sure whether that actually calls into the scandisk.exe file used in DOS mode though.

# ? Jun 28, 2016 16:41

Internet Janitor: May 17, 2008; "That isn't the appropriate trash receptacle."

Pollyanna posted:

Array languages like APL and K are also fascinating but I have no idea what I'd use them for Hard number crunching and data processing, right?

Actually at my new workplace K is used for a broad variety of things- number crunching is the core, but even stuff like web backends and automation scripts are mostly written in K. For those other tasks it doesn't shine as much, but it isn't really any worse than it would be to use JS or Python or any other conventional language.

# ? Jun 28, 2016 23:52

Hughlander: May 11, 2005

JawnV6 posted:

C's a lot of fun, but I don't really think of it as squeezing performance. More like making peek/poke machine accesses prettier.

Anyway, gcc 7.0.0 optimizations are breaking the Kernel build.

Which will now become the user's expectation and must not ever be moved.

To be fair, they introduced a bug in the kernel. They didn't break the build. If a GCC actually had a release that couldn't compile the kernel I'd be amazed at their release process not even doing a sanity check on one of the largest single use cases of it.

# ? Jun 29, 2016 21:50

krillin from Dragonball: Aug 8, 2013

Three days into my first coding job and I get to work on this: https://github.com/scoophealth/osca...A09Handler.java

Of note:

code:

private static void switchMatchingAppointment(LoggedInInfo loggedInInfo, String chartNo, List<Appointment> appointments) {
		// look through all appointments for matching demographic
		// set the here flag on matching
		// of not match throw exception.

		for (Appointment appointment : appointments) {
			logger.debug("checking appointment : " + appointment.getId());

			if (!isValidAppointmentStatusForMatch(appointment)) continue;

			if (chartNoMatches(loggedInInfo, appointment, chartNo)) {
				switchAppointmentStatus(appointment);
				return;
			}
		}

		throw (new IllegalStateException("Some one checking in who has no appointment."));
	}

	private static boolean isValidAppointmentStatusForMatch(Appointment appointment) {
		if ("H".equals(appointment.getStatus())) return (false);
		else if ("N".equals(appointment.getStatus())) return (false);
		else if ("C".equals(appointment.getStatus())) return (false);
		else if ("B".equals(appointment.getStatus())) return (false);

		return (true);
}

Use of continue, confusing function name, hard-coded values, and 5 separate returns? Great!

# ? Jun 29, 2016 23:30

Cuntpunch: Oct 3, 2003; A monkey in a long line of kings

I checked the history on this, just to see if MAYBE, MAYBE it had been slow degradation caused by changing requirements.

Nope, it got *written* like this originally and never touched since.

:smithicide:

code:

if( 	input.Foo != null ||
	input.Foo != null ||
	input.Foo != null ||
	input.Foo != null)

# ? Jun 30, 2016 00:11

qntm: Jun 17, 2009

What's the bet that getting input.Foo has side-effects?

# ? Jun 30, 2016 00:21

Adbot: ADBOT LOVES YOU

# ? Apr 23, 2024 11:56

JawnV6: Jul 4, 2004; So hot ...

Yeah, the quickest way to make that much worse is

code:

//  Avoids race condition

# ? Jun 30, 2016 00:26

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!

«‹›1501 »