My greatest coding sin is:

[chacham]

I've used goto

 33/3%

Not writing comments

 90/9%

Writing unnecessary comments

 18/2%

SQL with variables

 9/1%

Hardcoding passwords

 43/4%

Premature optimisation

 34/3%

Cargo cult programming

 9/1%

Brogramming

 2/0%

exit()ing instead of die()ing

 14/1%

Inadequate testing

 76/7%

Reinventing wheels

 79/8%

Other

 87/8%

I am a sin-free Saint

 549/53%

1043 total votes

[tye]

Copying and pasting code and then (usually only slightly) modifying the copy. So much effort and pain has resulted from that type of sin. Which means I've done it less than many, but I still have.

[Tux]

• Not changing sensible comment when changing the code that the comment was meant to so it was senseless and misleading after the change to the code.

[Old_Gray_Bear]

More than one of the above....

[chacham]

More than one of the above

More than one is your greatest?

[RichardK]

other :- not understanding the limitations of floating point!

[tye]

In my professional work, I've seen this particular sin most often in the form of "refusal to use floating point (due to lack of proper understanding of the limitations leading to a fear or dread or just undue reluctance to use floating point in so many situations where it would be better than some integer, usually in a poor choice of units)".

[Pope-O-Matik]

not understanding the limitations of floating point!

Just floating a point here: Floating points are not limited. Computer implemented floating point math, however, is.

[johngg]

I've confessed to (frequently) using goto ... in my defence, I didn't have much choice about it as I started out as a FORTRAN IV programmer. I haven't used one for decades since moving on to more modern languages :-/

[Anonymous Monk]

one-character variable names

[Happy-the-monk]

More than half of the above. Who can say what be the biggest of those sins?

Cheers, Sören

[dbuckhal]

Other: Being "too clever" with my code. Perl can promote fancy ways to do things, but keeping a (expression|statement) simple has its benefits.

[BrowserUk]

Spending too long trying to fix broken code rather than starting again from scratch.

Seven out of ten times when I start over I end up going back and fixing the original broken code anyway; but starting over from scratch, by removing the influence of the existing structure, variable names, comments etc., the realisation of what is wrong with the existing code often falls into place very quickly, where staring at the original code often leads one's brain down the same path that led to the original error in the first place.

And on the other three occasions, the new code comes together more quickly and cleanly because the second time around you've a clearer picture of what you need to achieve.

[hippo]

the second time around you've a clearer picture of what you need to achieve.

Undoubtedly. There's also the added benefit that you are coding to the spec as it stands now rather than to a bunch of feature-creeping incremental changes.

[Ratazong]

Indeed. Which reminds me of one of my all-time-favourite software-development related quotes:

Walking on water and developing software from a specification are easy if both are frozen. (Edward Berard)

[toolic]

Others:
• No usage documentation whatsoever.
• No version control.

[karlgoethebier]

Copying and pasting code and then (usually only slightly) modifying the copy... More than one of the above... (tye and Old_Gray_Bear)

I guess i'm in respectable society ;-)

And didn't the egyptian masons also just brick their dirt?

Best regards and a HNY, Karl

[BillKSmith]

What me a sinner? My variable names are not necessarily good, but at least they are long. I have used whatever it takes to avoid a simple goto. I have always tried to use a subroutine rather than repeating code (even when I know that some instances are likely to change in future versions). Why clutter my excellent code with comments? I am certainly open to reusing code. (To bad that I could never find exactly what I needed) My "clever" code makes it easy to see the bit picture at the expense of the details (or is it the other way around?) I read lots of programming books and manage to use every new trick I see. Most of my code is tested only with simple input. (Everyone knows that anything worth doing with a computer is to complicated to test more.)

[girarde]

I invented a text file format to store configuration items, either single items or lists, because I did not know what xml was at the time. It wasn't a freaking .ini file, though.

[karlgoethebier]

BTW, inadequate confessing of sins may cause trouble (Rome Episode from "Night on Earth" by Jim Jarmusch)

[anonymized user 468275]

To the self-appointed sin-free: Are you sure? I don't see how anyone can feasibly avoid re-inventing wheels unless you never achieve anything owing to spending all your time checking for originality. And that isn't the only trap for the self-righteous here. Commenting habits vary widely. Too widely to fit into the low standard deviation implied by the high "sin-free" percentage of voters. Oops!

Update: Let he who is without sin cast the first vote? or something like that.

[Lady_Aleena]

Other: I am guilty of more than one of the above. Anyone who knows me knows which ones.

[Anonymous Monk]

Clearly, many are either shameless liars or deluded.

[chacham]

Anyone else notice the node id for the first poll of '16, has both a 15 and a 16 in it?

[Anonymous Monk]

This poll should be multiple choice. I know that I have willfully committed at least half of these sins, and so am posting anonymously. I laugh at those who claim to be without sin.

[chacham]

The question is for the greatest. We have all committed multiple violations.

[Anonymous Monk]

That's apparently not true, read the results of your own poll. Over half of the responses claim to have committed no violations at all.