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!

«‹›1503 »

Tamba: Apr 5, 2010

http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/

quote:

I have never seen a computer system which handles names properly and doubt one exists, anywhere.

So, as a public service, I�m going to list assumptions your systems probably make about names. All of these assumptions are wrong. Try to make less of them next time you write a system which touches names.

*List of 40 items*

# ? Oct 25, 2012 11:53

Adbot: ADBOT LOVES YOU

# ? May 21, 2024 07:32

The Gripper: Sep 14, 2004; i am winner

My rule for names is that if I can't read your name enough to even attempt to pronounce it, then I don't care if my system breaks on it.

e: pretty much the same rules that would apply if I met someone in person.

# ? Oct 25, 2012 12:04

Malloc Voidstar: May 7, 2007; Fuck the cowboys. Unf. Fuck em hard.

The Gripper posted:

My rule for names is that if I can't read your name enough to even attempt to pronounce it, then I don't care if my system breaks on it.

Sure hope you're only developing programs to be used in extremely-localized areas, then. Supporting non-Latin alphabets is really easy. (and if you can't read Иван Иванович Иванов then that's your problem)

Similar article: Falsehoods programmers believe about time and part 2.

e: the SA forums are gross if you use unicode

Malloc Voidstar fucked around with this message at 12:27 on Oct 25, 2012

# ? Oct 25, 2012 12:24

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

Tamba posted:

http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/

Are you endorsing this article, or pointing it out for derision, or just presenting it as food for thought? Admittedly it might be very different to work on a system that's supposed to store and use people's real names (as opposed to their sign-in names for a service), but admonishing the assumption that people's names can be expressed in unicode, without giving suggestions as to a way to avoid making such assumptions, just reads like a load of self-important wank rather than serious discussion. I mean, the author points out a number of genuine pitfalls but I get the impression that he thinks that if someone comes along and tells you that their name is

, or

, your system should say "sure, no problem!" His list of peeves is of no use without some suggestion as to how to avoid them, given that some of them would be all but impossible practically to avoid in a working system.

The time articles linked by Aleksei have much the same issue. I would be interested in reading about why it is a misconception, as it is presented in part 2, that every year has either 365 or 366 days, but no explanatory link is provided.

# ? Oct 25, 2012 15:12

1337JiveTurkey: Feb 17, 2005

Hammerite posted:

The time articles linked by Aleksei have much the same issue. I would be interested in reading about why it is a misconception, as it is presented in part 2, that every year has either 365 or 366 days, but no explanatory link is provided.

The switch from julian to gregorian.

code:

$ cal 9 1752
   September 1752
Su Mo Tu We Th Fr Sa
       1  2 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30

# ? Oct 25, 2012 15:16

Malloc Voidstar: May 7, 2007; Fuck the cowboys. Unf. Fuck em hard.

Hammerite posted:

Are you endorsing this article, or pointing it out for derision, or just presenting it as food for thought? Admittedly it might be very different to work on a system that's supposed to store and use people's real names (as opposed to their sign-in names for a service), but admonishing the assumption that people's names can be expressed in unicode, without giving suggestions as to a way to avoid making such assumptions, just reads like a load of self-important wank rather than serious discussion. I mean, the author points out a number of genuine pitfalls but I get the impression that he thinks that if someone comes along and tells you that their name is , or , your system should say "sure, no problem!"

No:

quote:

All of these assumptions are wrong. Try to make less of them next time you write a system which touches names.

Pretty sure there are quite a few programmers who won't have paid any attention to names other than "it must accept a name like mine".
See sites that don't accept apostrophes or hyphens, or apply length requirements.

# ? Oct 25, 2012 15:24

Polio Vax Scene: Apr 5, 2009

Tamba posted:

http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/

Hi, my name is 37F89DA9-612C-498C-AFA5-AB9DF762E55F, but my friends call me 37.

# ? Oct 25, 2012 15:34

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 code:

define("TRUE", false);

if (TRUE) {
 echo "TRUE";
} else {
 echo "FALSE";
}

http://3v4l.org/Nc84q

# ? Oct 25, 2012 15:45

ToxicFrog: Apr 26, 2008

Hammerite posted:

Are you endorsing this article, or pointing it out for derision, or just presenting it as food for thought? Admittedly it might be very different to work on a system that's supposed to store and use people's real names (as opposed to their sign-in names for a service), but admonishing the assumption that people's names can be expressed in unicode, without giving suggestions as to a way to avoid making such assumptions, just reads like a load of self-important wank rather than serious discussion.

He is, in fact, talking about real names and not account names. A lot of systems expect your real name in addition to your account name (or will attempt to derive your account name from your real name, or will just assign you an account number or something). And a surprising number of those will choke and die if your name has an apostrophe in it, or a hyphen, or non-ASCII characters of any sort. Or it'll work fine, but changing your name later is either outright impossible or causes other things to break. And so on.

Yeah, it would be nice to have more information on how to avoid that issues, but it's a rant, not a lesson.

# ? Oct 25, 2012 16:12

Jewel: May 2, 2009

ToxicFrog posted:

He is, in fact, talking about real names and not account names. A lot of systems expect your real name in addition to your account name (or will attempt to derive your account name from your real name, or will just assign you an account number or something). And a surprising number of those will choke and die if your name has an apostrophe in it, or a hyphen, or non-ASCII characters of any sort. Or it'll work fine, but changing your name later is either outright impossible or causes other things to break. And so on.

Yeah, it would be nice to have more information on how to avoid that issues, but it's a rant, not a lesson.

If I legally get my name changed to Jewel I have a lot of problems too because sites like to say "OH NO YOUR NAME HAS 'JEW' IN IT, WE CAN'T ACCEPT THIS". A lot of sites currently do this even with my name as a Username, I've gotten banned from BigPond's gamearena TF2 servers because Jewel's an "offensive name" (I had to contact them to get it whitelisted, took a month), and I couldn't register to a few other services with the name either for the same reason. Never ever put a restriction on usernames, just let users report offensive ones if it's in an environment where it matters.

# ? Oct 25, 2012 16:26

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

Excuse me but I come from a long line of Bonerhitlers,

# ? Oct 25, 2012 16:43

The Gripper: Sep 14, 2004; i am winner

Plus there's weird barely-known countries like Australia, where if you're open and notorious with using an assumed name it's considered legal, and there's no formal process required to use it wherever you want. e; the only caveat is a few government agencies require written proof to issue you certain documents, but that's not necessary for the name to be legal.

If I openly called myself Jewbag McNegrofingers then I could have that name considered as my legal name, despite it being dumb and racisty.

I'm pretty sure you'd get no objections to banning me from your service for it though, since you're not obligated to provide me any services and I'd obviously be a worthless person.

The Gripper fucked around with this message at 16:47 on Oct 25, 2012

# ? Oct 25, 2012 16:44

tef: May 30, 2004; -> some l-system crap ->

Aleksei Vasiliev posted:

Similar article: Falsehoods programmers believe about time and part 2.

Similarly, http://qntm.org/gay

# ? Oct 25, 2012 17:01

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

tef posted:

Similarly, http://qntm.org/gay

There's a proposed genealogy data exchange format being developed by FamilySearch, a mormon company, and (surprisingly?) it will support gay marriages :allears:

_{The old standard (Gedcom) has hardcoded HUSB/WIFE fields; GedcomX Couple relationship types will allow any two persons.}

https://github.com/FamilySearch/gedcomx/blob/master/specifications/conceptual-model-specification.md

# ? Oct 25, 2012 17:54

darthbob88: Oct 13, 2011; YOSPOS

Manslaughter posted:

Hi, my name is 37F89DA9-612C-498C-AFA5-AB9DF762E55F, but my friends call me 37.

Any relation to Tom Lehrer's friend Hen3ry?

# ? Oct 25, 2012 18:00

Zamujasa: Oct 27, 2010; Bread Liar

Golbez posted:

PHP code:

define("TRUE", false);

if (TRUE) {
 echo "TRUE";
} else {
 echo "FALSE";
}

http://3v4l.org/Nc84q

Every time I try to justify to myself that PHP, while terrible, is at least perfectly usable if you're aware of some common problems, something like this comes along. It's not even that it's different. It's that it keeps changing back and forth for newer versions.

# ? Oct 25, 2012 18:11

Kal: Jun 3, 2007

code:

Main.userList.put(this.dm[1].toUpperCase().replaceAll("<", "").replaceAll(">", ""), new Friend(this.dm[1].toUpperCase().replaceAll("<", "").replaceAll(">", ""), this.dm[2].replaceAll("<", "").replaceAll(">", ""), this.s.getInetAddress(), this.dm[4].replaceAll("<", "").replaceAll(">", ""), this.now));

Found this gem is the source code of a chat server that is used for a college Java class I'm taking. And no, it's not written by some clueless student but by the person who ran the course last year :downs:

# ? Oct 25, 2012 18:37

Mogomra: Nov 5, 2005; simply having a wonderful time

Golbez posted:

PHP code:

define("TRUE", false);

if (TRUE) {
 echo "TRUE";
} else {
 echo "FALSE";
}

http://3v4l.org/Nc84q

http://3v4l.org/2s1sK

Seems like it might be a bug that's been introduced in 5.1.3. I don't see why anyone would ever want to do stuff like that in PHP unless it was a practical joke, though. There's always the possibility that somewhere deep in the bowels of the PHP documentation, there's a note saying you can't actually define constants using reserved words too. "Nope, guys. Working as intended!" - :shepface:

Either way, you can't really call it a coding horror if you're trying to alter something like the truthiness of TRUE at runtime, and you get unexpected results. That kind of nonsense might have worked in days gone by, but we now we have evolved sensibilities, man!

I'm thinking that TRUE and FALSE aren't actually constants, and calling "define("TRUE", $some_val);" creates a constant that will always be trumped by TRUE in the interpreter.

_{_{_{Note: I'm not defending PHP in general, I'm defending it in this specific case. There are legitimate horrors in that language, but we can't go around calling things horrors just because they don't behave the way we expect them to.}}}

# ? Oct 25, 2012 18:40

nielsm: Jun 1, 2009

I believe "true" is a keyword in PHP, and keywords are case insensitive. So in version 5.1.3 defines might have been made to resolve after keywords are checked for.

# ? Oct 25, 2012 19:11

Zombywuf: Mar 29, 2008

Aleksei Vasiliev posted:

No:
Pretty sure there are quite a few programmers who won't have paid any attention to names other than "it must accept a name like mine".
See sites that don't accept apostrophes or hyphens, or apply length requirements.

I feel it is wrong to call these people programmers. Although there is also a very good chance they are programmers but their boss is a micromanaging idiot. However the unicode thing is not an assumption, it is the point where you throw up your hands and go "look, you're just going to have to meet me in the middle here, wtf do I put in the email template 'Dear _____,' and no; I'm not putting an image there."

# ? Oct 25, 2012 19:41

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

Golbez posted:

PHP code:

define("TRUE", false);

if (TRUE) {
 echo "TRUE";
} else {
 echo "FALSE";
}

http://3v4l.org/Nc84q

Stuff like this is going to happen as PHP devs try to make PHP behave more consistently.

Here is the entry for it in the changelog for 5.1.3

quote:

* Eliminated run-time constant fetching for TRUE, FALSE and NULL

I do not see this as a Coding Horror, more like it's trying to correct a previous Coding Horror. Considering this change was made to PHP more than 6 years ago and it's doing something that no sane coder would do, I don't think it's that big of a deal.

# ? Oct 25, 2012 19:42

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

Zamujasa posted:

... It's that it keeps changing back and forth for newer versions.

According to the linked page it doesn't? It gives one result for versions 5.1.2 and earlier, another result for versions 5.1.3 and later

# ? Oct 25, 2012 19:55

ohgodwhat: Aug 6, 2005

Carthag posted:

There's a proposed genealogy data exchange format being developed by FamilySearch, a mormon company, and (surprisingly?) it will support gay marriages

_{The old standard (Gedcom) has hardcoded HUSB/WIFE fields; GedcomX Couple relationship types will allow any two persons.}

https://github.com/FamilySearch/gedcomx/blob/master/specifications/conceptual-model-specification.md

You'd think they'd generalize it to handle polygamy?

# ? Oct 25, 2012 20:10

bobthecheese: Jun 7, 2006; Although I've never met Martha Stewart, I'll probably never birth her child.

nielsm posted:

I believe "true" is a keyword in PHP, and keywords are case insensitive. So in version 5.1.3 defines might have been made to resolve after keywords are checked for.

No, if you try to

code:

define('true', false);

then it'll give you a notice telling you that 'true' is already defined.

So the horror here is that it's allowing you to make a variable definition, not notifying you that it won't work, and then ignoring it. The example itself was stupid (redefining true as false), but I can think of a semi legitimate reason to make a define that would say, define TRUE as a string 'true', or as 1, or as a number of things to deal with things like input parsing.

I'm not saying that it's a good thing to do, but I'm sure that someone, somewhere, would think that it's a good idea, and when that happens, PHP should either let them do it, or be consistent with notifying them that it won't work.

# ? Oct 25, 2012 20:15

Janitor Prime: Jan 22, 2004; PC LOAD LETTER

What da fuck does that mean; Fun Shoe

Horrors all the way down
https://crypto.stanford.edu/~dabo/pubs/abstracts/ssl-client-bugs.html

:stare:

# ? Oct 25, 2012 20:17

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

ohgodwhat posted:

You'd think they'd generalize it to handle polygamy?

Yeah actually that'd make sense. Though I guess they'd just model a polygamous person by having them in several concurrent marriages. Depends on the exact nature of the marriage I guess, I mean if it's an "equal" one where everybody is married to everybody, or it's one dude with 5 wives.

I should open an issue on github :haw:

# ? Oct 25, 2012 20:22

qntm: Jun 17, 2009

The Gripper posted:

My rule for names is that if I can't read your name enough to even attempt to pronounce it, then I don't care if my system breaks on it.

e: pretty much the same rules that would apply if I met someone in person.

Not a fan of the Chinese, huh

# ? Oct 25, 2012 22:32

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

The Gripper posted:

My rule for names is that if I can't read your name enough to even attempt to pronounce it, then I don't care if my system breaks on it.

e: pretty much the same rules that would apply if I met someone in person.

How often are you reading someone's name in person? Don't you just repeat their name after they introduce themselves? Do you live in a conference centre?

# ? Oct 25, 2012 23:48

The Gripper: Sep 14, 2004; i am winner

I spend most of my free time at AA, reading nametags.

# ? Oct 25, 2012 23:51

DaTroof: Nov 16, 2000; CC LIMERICK CONTEST GRAND CHAMPION
There once was a poster named Troof
Who was getting quite long in the toof

The Gripper posted:

I spend most of my free time at AA, reading nametags.

And I'm sure the frequency of members using unpronounceable symbols to identify themselves is a problem that destroys meetings.

# ? Oct 26, 2012 01:50

Opinion Haver: Apr 9, 2007

Carthag posted:

Yeah actually that'd make sense. Though I guess they'd just model a polygamous person by having them in several concurrent marriages. Depends on the exact nature of the marriage I guess, I mean if it's an "equal" one where everybody is married to everybody, or it's one dude with 5 wives.

I should open an issue on github

I can't wait until people have to learn graph theory to understand marriages.

# ? Oct 26, 2012 02:40

God of Mischief: Oct 22, 2010

yaoi prophet posted:

I can't wait until people have to learn graph theory to understand marriages.

Is that when the Deep South allows homosexual polygamy?

"Well, I'm married to my grandma, nephew, sister, father and myself (and my pig but he don't count)."

# ? Oct 26, 2012 04:36

Deus Rex: Mar 5, 2005

yaoi prophet posted:

I can't wait until people have to learn graph theory to understand marriages.

classic: http://stackoverflow.com/questions/6163683/cycles-in-family-tree-software

# ? Oct 26, 2012 04:48

Tern: Oct 18, 2012

God of Mischief posted:

Is that when the Deep South allows homosexual polygamy?

"Well, I'm married to my grandma, nephew, sister, father and myself (and my pig but he don't count)."

I suppose then the rule will be that one must be able to construct a spanning tree for each family.

Edit: crap, Deus Rex kinda beat me to it.

# ? Oct 26, 2012 04:50

fritz: Jul 26, 2003

1337JiveTurkey posted:

The switch from julian to gregorian.

code:

$ cal 9 1752
   September 1752
Su Mo Tu We Th Fr Sa
       1  2 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30

That was only the switch in the UK & subsidiaries, and I think they were late because the Gregorian calendar was papist. (and didn't places Greece switch really late, like early 20th century?)

Otto Skorzeny posted:

The idea that + should be commutative is not remotely unreasonable

It's clearly not always, though. Floating point isn't guaranteed to be commutative.

# ? Oct 26, 2012 04:53

hobbesmaster: Jan 28, 2008

fritz posted:

It's clearly not always, though. Floating point isn't guaranteed to be commutative.

Floating point underflow and related issues are to be expected though.

# ? Oct 26, 2012 05:07

Jabor: Jul 16, 2010; #1 Loser at SpaceChem

Floating point + clearly isn't associative, but I can't think of anything off the top of my head where it's not commutative.

# ? Oct 26, 2012 05:08

tef: May 30, 2004; -> some l-system crap ->

fritz posted:

That was only the switch in the UK & subsidiaries, and I think they were late because the Gregorian calendar was papist. (and didn't places Greece switch really late, like early 20th century?)

http://en.wikipedia.org/wiki/February_30#Swedish_calendar the swedish changeover was the best

# ? Oct 26, 2012 05:11

Outer Science: Dec 21, 2008; Daisangen

Jabor posted:

Floating point + clearly isn't associative, but I can't think of anything off the top of my head where it's not commutative.

String concatenation in languages that use it as such, for one obvious (but vaguely cheaty :v:

) answer

# ? Oct 26, 2012 05:31

Adbot: ADBOT LOVES YOU

# ? May 21, 2024 07:32

Jewel: May 2, 2009

Ahaha, I found some good addon for Chrome that I wanted to use, but it had a lot of features that broke some stuff, where I only wanted one of the features, so I downloaded the .crx and extracted it with winrar, giving me access to all the code so I could just change it and put it back into extensions. This line of code is in one of the files, I have no idea what it does:

JavaScript code:

function bug2(what) {
	if (1==2) { console.log('>>>'+what); }
}

It used like this about 5 times throughout the 300 line script:

JavaScript code:

bug2('this page is good? '+ good +' ('+ taburl +')' );

But since there's the "if (1==2)" won't that never run??

Edit: Oh, right, debugging exists, could change it to 1==1 whenever they wanted to debug. Carry on!!

Jewel fucked around with this message at 05:39 on Oct 26, 2012

# ? Oct 26, 2012 05:34

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!

«‹›1503 »