How Fast Can You Debug?

August 21, 2014 10:12 AM

http://www.codefights.com/
posted by boo_radley (36 comments total)

I get only a login screen. Is there more to this?
posted by Philosopher Dirtbike at 10:14 AM on August 21, 2014 [1 favorite]

I got little code snippets with errors to correct.
posted by kmz at 10:19 AM on August 21, 2014

I get repeated feelings of failure.
posted by clawsoon at 10:22 AM on August 21, 2014 [9 favorites]

Soon to be an actual performance assessment tool, I wonder?
posted by clockzero at 10:25 AM on August 21, 2014 [1 favorite]

I get only a login screen. Is there more to this?

Do you have Ghostery or similar extension running? I had to switch it off (or use another browser) to make the game visible.

While I'm here let me say for the record that I SUCK at debugging!
posted by jeremias at 10:26 AM on August 21, 2014

clockzero: "Soon to be an actual performance assessment tool, I wonder?"

I wish. The mechanical FIND BUG FAST assessors I've used are completely brutal in comparison to this.
posted by boo_radley at 10:32 AM on August 21, 2014

Soon to be an actual performance assessment tool, I wonder?

For coders, tests will replace resumes and interviews
posted by fatehunter at 10:35 AM on August 21, 2014

Doesn't work very well on phones. If you accidentally hit return while you're hunting for backslash on the crazy keyboard then you can't delete the indentation it automatically adds!

Otherwise pretty neat. Reminds me of that coding roguelike we saw a few months back.
posted by wilberforce at 11:02 AM on August 21, 2014

The answer is none. None fast can I debug.
posted by shmegegge at 11:17 AM on August 21, 2014 [6 favorites]

Not sure how well this works. One chunk of code that had four lines in it could be fixed with a ternary to reduce it to one bug-free statement, but the app complains about changing the number of lines, sadly.
posted by Mr. Six at 11:29 AM on August 21, 2014 [2 favorites]

Apparently it would help if I understood Javascript syntax. Not doing too badly regardless though.
posted by pharm at 11:35 AM on August 21, 2014 [1 favorite]

Mr. Six: "Not sure how well this works. One chunk of code that had four lines in it could be fixed with a ternary to reduce it to one bug-free statement, but the app complains about changing the number of lines, sadly."

bug fixing is not refactoring, although I support your thesis.
posted by boo_radley at 11:39 AM on August 21, 2014 [3 favorites]

I think I did okay, probably because I've been working in JavaScript a lot recently. I ended up with 350 points before I ran out of time.
posted by zixyer at 11:54 AM on August 21, 2014

I did much better once I realized it was JavaScript and not Java. I guess that explains my problems with my second marriage too.
posted by ubiquity at 12:18 PM on August 21, 2014 [9 favorites]

I don't want people whose goal is speed doing my debugging, thanks
posted by thelonius at 12:29 PM on August 21, 2014 [3 favorites]

Apparently I rule at catching 'assignment vs equality' bugs, suck at everything else though!
posted by TwoWordReview at 12:34 PM on August 21, 2014

I don't think performance on these problems says much about your capacity as a software developer. These aren't the kind of bugs I really get hung up on.
posted by fivebells at 1:34 PM on August 21, 2014 [2 favorites]

Mr. Six: "Not sure how well this works. One chunk of code that had four lines in it could be fixed with a ternary to reduce it to one bug-free statement, but the app complains about changing the number of lines, sadly."

Well it does say at the beginning: "There is one bug on one line of the below code." (emphasis mine)
posted by thewalledcity at 1:38 PM on August 21, 2014

I think their game has a bug in it; I corrected the first snippet, then got the corrected version of the snippet as my second test and couldn't continue. Ironic.
posted by axiom at 1:39 PM on August 21, 2014 [2 favorites]

I've played it a few more times and noticed that they reuse the same snippets with different bugs each time. You probably didn't notice the bug the second time around.
posted by zixyer at 1:43 PM on August 21, 2014

I've played it a few more times and noticed that they reuse the same snippets with different bugs each time. You probably didn't notice the bug the second time around.

Yeah, I thought the same thing as axiom once but noticed they actually changed something else about the code.
posted by kmz at 2:00 PM on August 21, 2014

That was quite fun. I did keep getting distracted, though, by wanting to correct the spelling of "concatenating" in one of the comments, but not being able to because of the "only change one line" rule.
posted by Law of Demeter at 2:10 PM on August 21, 2014

Is it supposed to show the correct answer if you don't get it right? Because mine isn't.
posted by ob1quixote at 4:00 PM on August 21, 2014

This is a neat idea. Anybody know if there are games(?)/tools(?) like this for other programming languages?
posted by Flipping_Hades_Terwilliger at 4:39 PM on August 21, 2014

This is a neat idea. Anybody know if there are games(?)/tools(?) like this for other programming languages?

Where I'm from we call it "work".
posted by mistersquid at 5:06 PM on August 21, 2014 [7 favorites]

Debugging hundreds of thousands of lines of transliterated JavaScript is a much different experience, I will tell you.
posted by RobotVoodooPower at 10:51 PM on August 21, 2014

I found a huge bug in all of these. They're written JavaScript.
posted by qxntpqbbbqxl at 11:36 PM on August 21, 2014 [4 favorites]

I don't debug, I rebug :(
posted by comealongpole at 2:48 AM on August 22, 2014 [2 favorites]

I played this once before, and got like 90. This time I got around 300, but I got held up on a "bug" where the line of code they wanted me to fix wasn't the only problem with the function. I had to stare at the code and intuit "oh, that's the line they want me to fix" instead of fixing the actual bug, which was frustrating.

Also, the timer means you can't look up some behavior (I always need to reference how regular expressions work in any language, because they're subtly different in every one I use).
posted by jamuraa at 6:05 AM on August 22, 2014

I am finding it hard to resist the temptation to just fix one line of code. I mean, the problem with:

/**
 * @Input: a string
 * @Output: true if @input has length 3, false otherwise
 */
function checkData(string) {
  if (string.length == 3) {
    return true;
  } else {
    return string.length;
  }
}

is not the final return, it's that they didn't just return the comparison result. As in:

/**
 * @Input: a string
 * @Output: true if @input has length 3, false otherwise
 */
function checkData(string) {
  return (string.length == 3);
}

posted by pulposus at 7:28 AM on August 22, 2014

Do you have to sign in to play this, or am I just less-than-stellar at following instructions this morning?
posted by Pudhoho at 8:50 AM on August 22, 2014

No sign-in required, as near as I can tell. (Unless you want to compare scores with social-media friends?)
posted by Mr. Six at 9:45 AM on August 22, 2014

Pudhoho: “Do you have to sign in to play this, or am I just less-than-stellar at following instructions this morning?”

You don't but you have to allow Javascript from codefights.com and cdn.mxpnl.com as well as enable MixPanel for the domain in Ghostery.
posted by ob1quixote at 11:47 AM on August 22, 2014

Thank you, Mr. Six and ob1quixote! I forgot about Ghostery. I feel like I put my shorts on backwards.
posted by Pudhoho at 12:42 PM on August 22, 2014

Ugh the style and lack of optimization. No wonder they can't find their bugs themselves.
posted by captaincrouton at 12:43 PM on August 22, 2014 [2 favorites]

Yeah, the hardest part of this is not simply replacing the whole function with a unambiguously correct one-or-two-line version. Writing a for-loop to do a base-10 logarithm, for shame!
posted by Gilead at 5:48 PM on August 23, 2014 [1 favorite]