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 »

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

Manslaughter posted:

I've been given an assignment to create an XSL that will insert the appropriate RTF tags to format XML to a client's liking.

You heard right, RTF. I'm going to manually insert these tags all over an XSL file so that XML goes in one end and a pretty doc (with an RTF extension) comes out the other.

Will post trip report if my brain does not melt out one of my ears before I'm done.

bonus round: One of the items on the client list is a watermark

Did you have a discussion with the source of the request to see why they want/need that? 90% of the time when a totally nonsensical request comes through, it's because no one understands what they actually want or need.

# ? Feb 3, 2014 22:40

Adbot: ADBOT LOVES YOU

# ? May 17, 2024 16:19

Munkeymon: Aug 14, 2003; Motherfucker's got an
armor-piercing crowbar! Rigoddamndicu𝜆ous.

Plorkyeran posted:

Might as well save some time and annoyance and just kill yourself now.

The note better be well-formed or the attempt will probably just error out.

# ? Feb 3, 2014 22:50

Polio Vax Scene: Apr 5, 2009

Ithaqua posted:

Did you have a discussion with the source of the request to see why they want/need that? 90% of the time when a totally nonsensical request comes through, it's because no one understands what they actually want or need.

Client needs: Docs that are formatted to their preferences
Our needs: Money, and a new system that supports something other than RTF (yeah right)

TBH this client has no idea what they want. They'll request customization, get billed for it, then request that exact same customization be removed (and you better believe they get billed the same amount for removal).

# ? Feb 3, 2014 23:13

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

Manslaughter posted:

I've been given an assignment to create an XSL that will insert the appropriate RTF tags to format XML to a client's liking.

You heard right, RTF. I'm going to manually insert these tags all over an XSL file so that XML goes in one end and a pretty doc (with an RTF extension) comes out the other.

Will post trip report if my brain does not melt out one of my ears before I'm done.

bonus round: One of the items on the client list is a watermark

At first I read that as "XLS", and thought they wanted Excel macros. So, look on the bright side...

# ? Feb 3, 2014 23:52

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

Suspicious Dish posted:

At first I read that as "XLS", and thought they wanted Excel macros. So, look on the bright side...

I'm not sure that'd be worse. If you have access to Word's stuff from excel macros then it'd actually be easier to create a RTF file from there than via XSLT.

# ? Feb 3, 2014 23:56

shrughes: Oct 11, 2008; (call/cc call/cc)

Steve French posted:

If you're going to build a shed with pointers,

C code:

int isPalindrome(char *s, size_t len) {
  char *e = s + len - 1;
  while (s < e) if (*s++ != *e--) return 0;
  return 1;
}

That is indeed an appropriate contribution for this thread, since it has undefined behavior on one of its inputs.

# ? Feb 4, 2014 00:10

Harik: Sep 9, 2001; From the hard streets of Moscow
First dog to touch the stars; Plaster Town Cop

Jewel posted:

I was confused about why "warning = limit * persentage" wouldn't work. I haven't worked with C++ in a little while so I thought "hm, does 'limit * persentage' create a new unsigned char of the multiplication, then cast that to long? I thought that makes sense so I did a quick test to see what the output would be. I got:
C++ code:
#include <iostream>

void multiply(unsigned char a, unsigned char b)
{
    unsigned long result;
    result = a * b;

    std::cout << +a << " * " << +b << " = " << result << std::endl;
}

int main()
{
    multiply(5, 5);
    multiply(255, 255);

    return 0;
}
With the output of:
code:
5 * 5 = 25
255 * 255 = 65025
Why is this? It seems to work? So my two questions: Why did their code snippet not work, and why does this work with regards to casting and overflow?

It's on an 8-bit processor, where anything other than 8bit * 8bit multiply requires a function call. Even if C didn't specify promotion, your compiler would have to go out of it's way to poo poo on you since you're building on a machine with at least 32-bit registers if not 64, so it'd have to multiply & mask. As far as the original horror, on the hardware it should be u8*u8=u16 (two register destination). So probably "broken compiler", it wouldn't be the first time. Casting or assigning either operand to a long fixed it though:

C++ code:

void multiply(unsigned char a, unsigned char b) {
    unsigned long result=a;

    result = result * b;
}

The uninitialized variable in the example was me, I'm the cut&paste horror when distilling a monster 200 line function down to the fuckup.

I keep "persentage" around to remind me how much I hate cleaning up other people's messes. It's a great source of rage when I have to poke around the spaghetti pile for whatever minor misfeature I need to add.

Except I don't anymore, since it's all rebuilt on a new chip so oh well that PIC-specific code just gosh darn didn't port and I had to re-implement.

# ? Feb 4, 2014 00:45

Simulated: Sep 28, 2001; Lowtax giveth, and Lowtax taketh away.; College Slice

Scaevolus posted:

Fizzshedding: dozens of replies arguing about implementations of a trivial interview question.

This happens every time someone posts about interview questions. Some of you nerds just can't help making it into an e-penis contest instead of focusing on, you know, the actual discussion. Hey everyone, look at meeeeeeeeee!

# ? Feb 4, 2014 03:46

Smugdog Millionaire: Sep 14, 2002; 8) Blame Icefrog

Many programmers lack the self-awareness necessary to understand when writing code is not the solution, and the code they write winds up in this thread.

:ironicat:

# ? Feb 4, 2014 07:49

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

Smugdog Millionaire posted:

Many programmers lack the self-awareness necessary to understand when writing code is not the solution, and the code they write winds up in this thread.

Really, we're just being efficient.

# ? Feb 4, 2014 08:40

shrughes: Oct 11, 2008; (call/cc call/cc)

Ender.uNF posted:

This happens every time someone posts about interview questions. Some of you nerds just can't help making it into an e-penis contest instead of focusing on, you know, the actual discussion. Hey everyone, look at meeeeeeeeee!

The discussion is whatever we make the discussion.

# ? Feb 4, 2014 09:17

Steve French: Sep 8, 2003

Ender.uNF posted:

This happens every time someone posts about interview questions. Some of you nerds just can't help making it into an e-penis contest instead of focusing on, you know, the actual discussion. Hey everyone, look at meeeeeeeeee!

Wait, so are you participating in the actual discussion?? Here I thought it was reasonable to talk about code.

I helped start the "derail" by asking some pointed questions intended to highlight the possibility that someone had misunderstood why their interview performance was perceived as poor.

THEN I put forth my entry in the e-penis contest because guess what? I actually enjoy thinking of and implementing various solutions to these sorts of problems.

AND I ended up learning something from it. From shrughes, who you all seem to hate for some reason or another. Maybe be less upset about things I guess?

# ? Feb 4, 2014 17:50

return0: Apr 11, 2007

shrughes posted:

That is indeed an appropriate contribution for this thread, since it has undefined behavior on one of its inputs.

Out of curiosity what's the UB?

# ? Feb 4, 2014 18:16

coffeetable: Feb 5, 2006; TELL ME AGAIN HOW GREAT BRITAIN WOULD BE IF IT WAS RULED BY THE MERCILESS JACKBOOT OF PRINCE CHARLES

YES I DO TALK TO PLANTS ACTUALLY

return0 posted:

Out of curiosity what's the UB?

I've written all of ~ten lines of C in my life, but undefined behaviour is usually at the boundaries. In this case, set len = 0. Then e = s - 1, and (if I'm remembering rightly) since pointers to arrays are pointers to their first element, e is a pointer to one before the first element, which is an invalid pointer.

# ? Feb 4, 2014 18:23

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

Obviously it just requires that you start your strings at the second byte of the allocation.

# ? Feb 4, 2014 18:33

Admiral H. Curtiss: May 11, 2010; I think there are a bunch of people who can create trailing images. I know some who could do this as if they were just going out for a stroll.

Isn't that a moot point anyway? At len=0, the while loop ends up as (s < s-1), which immediately fails and the function returns 1. You're never dereferencing the pointer that way.

# ? Feb 4, 2014 19:20

qntm: Jun 17, 2009

code:

if (capturedXML.startsWith("HEX")) {
    ...
}

# ? Feb 4, 2014 19:23

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

Admiral H. Curtiss posted:

Isn't that a moot point anyway? At len=0, the while loop ends up as (s < s-1), which immediately fails and the function returns 1. You're never dereferencing the pointer that way.

I was a bit surprised, but I checked the C99 draft and it says that only address calculations within an array and just past the end are defined. An address just before the beginning is right out, regardless of whether you dereference it.

# ? Feb 4, 2014 19:27

qntm: Jun 17, 2009

Gazpacho posted:

I was a bit surprised, but I checked the C99 draft and it says that only address calculations within an array and just past the end are defined. An address just before the beginning is right out, regardless of whether you dereference it.

There aren't any arrays in that code...?

# ? Feb 4, 2014 19:30

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

Generally speaking, if the pointer isn't pointing into an array then you've done something undefined before you call the function.

# ? Feb 4, 2014 19:32

coffeetable: Feb 5, 2006; TELL ME AGAIN HOW GREAT BRITAIN WOULD BE IF IT WAS RULED BY THE MERCILESS JACKBOOT OF PRINCE CHARLES

YES I DO TALK TO PLANTS ACTUALLY

qntm posted:

There aren't any arrays in that code...?

She's passing a pointer and a length. That's an idiom for passing an array, and in this case it's an array of chars, which is one way to represent a string.

coffeetable fucked around with this message at 20:07 on Feb 4, 2014

# ? Feb 4, 2014 20:04

Fuck them: Jan 21, 2011; and their bullshit

The reason I thought "Oh, a higher level language - I'll split and flip one half then compare!" was because I had just come from C.

Welp.

# ? Feb 4, 2014 20:25

Bonfire Lit: Jul 9, 2008; If you're one of the sinners who caused this please unfriend me now.

qntm posted:

There aren't any arrays in that code...?

For arithmetic operations, pointers to objects that are not part of an array shall behave the same as pointers to the first element of a one-element array of the same object type. As soon as your calculation produces a pointer that lies outside of that (real or fictional) array, even as a temporary value, the behavior is undefined - unless the pointer points one past the last element, which is allowed as a special exception.

# ? Feb 4, 2014 22:05

Fuck them: Jan 21, 2011; and their bullshit

edit: wrong thread

Fuck them fucked around with this message at 22:33 on Feb 4, 2014

# ? Feb 4, 2014 22:29

shrughes: Oct 11, 2008; (call/cc call/cc)

Admiral H. Curtiss posted:

Isn't that a moot point anyway? At len=0, the while loop ends up as (s < s-1), which immediately fails and the function returns 1. You're never dereferencing the pointer that way.

On a somewhat plausible architecture, s could underflow resulting in an incorrect result from the comparison. On other architectures, if the compiler inlined the function, the compiler could hypothetically assume that len > 0 based on the code it sees (because undefined behavior) and generate assembly code that doesn't do the right comparison the first time it enters the loop.

# ? Feb 4, 2014 23:23

Tesseraction: Apr 5, 2009

That actually sounds pretty interesting - which compilers do that? I'm self-studying optimisation at the moment so that'd be a good thing to look at.

# ? Feb 4, 2014 23:44

shrughes: Oct 11, 2008; (call/cc call/cc)

I'm not naming a specific optimization that a compiler might do or any compiler behavior I specifically know about. This is just a hypothetical code generation decision (which seems pretty unlikely).

You could imagine it generating some assembly that for whatever reason looks something like this:

code:


    e = s + len - 1
    cmp s, e
    jz bottom_of_loop
top_of_loop:
    ...

    incr s
    decr e
    cmp s, e
    jl top_of_loop
end_of_loop:

where it decides without any particular optimization in mind that it would like to use a jump-if-zero instruction instead of jump-if-not-less-than because it "knows" len is not 0. And where it has comparisons at both ends of the loop (which happens sometimes).

You can see other circumstances where undefined behavior results in a surprise optimization, like seen http://lwn.net/Articles/342330/ <- there.

# ? Feb 5, 2014 00:16

return0: Apr 11, 2007

shrughes posted:

You can see other circumstances where undefined behavior results in a surprise optimization, like seen http://lwn.net/Articles/342330/ <- there.

This is pretty cool tbh

# ? Feb 5, 2014 01:07

Simulated: Sep 28, 2001; Lowtax giveth, and Lowtax taketh away.; College Slice

Steve French posted:

Wait, so are you participating in the actual discussion?? Here I thought it was reasonable to talk about code.

I helped start the "derail" by asking some pointed questions intended to highlight the possibility that someone had misunderstood why their interview performance was perceived as poor.

THEN I put forth my entry in the e-penis contest because guess what? I actually enjoy thinking of and implementing various solutions to these sorts of problems.

AND I ended up learning something from it. From shrughes, who you all seem to hate for some reason or another. Maybe be less upset about things I guess?

# ? Feb 5, 2014 01:20

Tesseraction: Apr 5, 2009

shrughes posted:

where it decides without any particular optimization in mind that it would like to use a jump-if-zero instruction instead of jump-if-not-less-than because it "knows" len is not 0. And where it has comparisons at both ends of the loop (which happens sometimes).

You can see other circumstances where undefined behavior results in a surprise optimization, like seen http://lwn.net/Articles/342330/ <- there.

Interesting - thanks for the clarification. It's a good example of a compiler not following its supposed principles (i.e. not changing semantics) due to ambiguous syntax. Good eye!

# ? Feb 5, 2014 01:28

coffeetable: Feb 5, 2006; TELL ME AGAIN HOW GREAT BRITAIN WOULD BE IF IT WAS RULED BY THE MERCILESS JACKBOOT OF PRINCE CHARLES

YES I DO TALK TO PLANTS ACTUALLY

Steve French posted:

AND I ended up learning something from it. From shrughes, who you all seem to hate for some reason or another.

It's possible for someone to be both very knowledgeable and a bit of an rear end in a top hat.

# ? Feb 5, 2014 01:29

Tesseraction: Apr 5, 2009

Yeah I mean I disagree with shrughes on most social points but he knows his C (family).

# ? Feb 5, 2014 01:31

FamDav: Mar 29, 2008

2banks1swap.avi posted:

The reason I thought "Oh, a higher level language - I'll split and flip one half then compare!" was because I had just come from C.

Welp.

Why?

# ? Feb 5, 2014 01:32

shrughes: Oct 11, 2008; (call/cc call/cc)

Tesseraction posted:

Yeah I mean I disagree with shrughes on most social points

I doubt this is true. I only argue with irrational arguments people here make because they're irrational, not because the end conclusion is right or wrong.

# ? Feb 5, 2014 01:34

Tesseraction: Apr 5, 2009

shrughes posted:

I doubt this is true. I only argue with irrational arguments people here make because they're irrational, not because the end conclusion is right or wrong.

I will willingly and happily argue this, hoping to be wrong, but not in this thread.

# ? Feb 5, 2014 01:38

Deus Rex: Mar 5, 2005

Ender.uNF posted:

# ? Feb 5, 2014 01:49

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

Tesseraction posted:

Interesting - thanks for the clarification. It's a good example of a compiler not following its supposed principles (i.e. not changing semantics) due to ambiguous syntax. Good eye!

I think you still might not fully understand what undefined behavior is all about.

# ? Feb 5, 2014 02:00

Tesseraction: Apr 5, 2009

No, I'm pretty sure I do. Sorry my post produced undefined behaviour in your interpretation?

# ? Feb 5, 2014 02:34

shrughes: Oct 11, 2008; (call/cc call/cc)

Tesseraction posted:

No, I'm pretty sure I do. Sorry my post produced undefined behaviour in your interpretation?

You said the compiler changed semantics when in fact it didn't.

# ? Feb 5, 2014 02:55

Adbot: ADBOT LOVES YOU

# ? May 17, 2024 16:19

Sedro: Dec 31, 2008

FamDav posted:

Why?

It would work half the time

# ? Feb 5, 2014 02:57

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 »