• Panda (unregistered)

    Frist again?

  • Anomaly (unregistered) in reply to Panda
    Panda:
    Frist again?

    Technically but it doesn't count on a double article day like today. Only Frist Prime counts.

  • o11c (unregistered)

    Now that the contest is over, I'll link to mine (not putting it inline, that would kill the syntax highlighting) https://gist.github.com/o11c/5667876

    I would've submitted except for the "have to download and run this product" thing.

  • LK (unregistered) in reply to Anomaly
    Panda:
    Frist again?

    First is so passé, says today's XKCD: http://xkcd.com/1258/

    But apparently Frist still lives on in some WTF'y corners of the web.

  • (cs)

    I considered entering. I had a very cobbled together solution that was meant to seem like it had been developed by multiple people, each using their own personal favorite language regardless of appropriateness. Most importantly, it's "secure".

    The idea was simple. First, a NSIS wizard asks a few questions, including username and password. It parses the answers into a float (dev likes floats because they're "more versatile than integers"), converts that float to a string (second dev prefers strings because they're "even more versatile"), appends a random 3-bit salt (which, of course, is written as a number 0-7), RSA encrypts that string using the password as a key, and prepends the md5 hash of the username. The result is saved to a plaintext file called "C:\secure\answer.txt"

    Then, to actually get your answer, you have to open answer.txt up in a separate program, written in AutoIt 3. It'll ask for your username, and if the md5 of that username matches the first characters of the file, it'll try to decrypt the string using the password. Once the string is decrypted, it throws out the first character, separates the remaining string out to individual digits, which correspond to the answers you gave to the questions it asks, and makes one of two decisions based on those answers.

    For the record, the decision is meant to actually correspond to the questions asked. In theory, it's supposed to give an answer that's actually right for the choices you made in the wizard (let's say it's for budgeting, and the question is whether you can afford something, it asks you a few questions about your finances and the short- and long-term costs of the product, and comes up with either "buy" or "don't buy"). However, the result is completely random because of a bug: we left the salt in (which we'd appended), and threw away the first character by mistake.

    In the end, I didn't feel like coding this. NSIS and AutoIt 3 are both incredibly easy, but I was incredibly lazy. Ah, well, I certainly couldn't compete with the likes of some of these runners up, anyway.

  • Meep (unregistered)

    Nothing I could write could compare to the code I see at work.

  • (cs) in reply to Meep
    Meep:
    Nothing I could write could compare to the code I see at work.

    Pictures or it didn't happen.

  • (cs) in reply to Anomaly
    Anomaly:
    Panda:
    Frist again?

    Technically but it doesn't count on a double article day like today. Only Frist Prime counts.

    FRIST BE THAT

  • Meep (unregistered) in reply to Matt Westwood
    Matt Westwood:
    Meep:
    Nothing I could write could compare to the code I see at work.

    Pictures or it didn't happen.

    The worst stuff is hidden away on a client system, for example, thousands of lines of awk that is living inside a series of giant bash scripts.

  • (cs) in reply to Meep
    Meep:
    The worst stuff is hidden away on a client system, for example, thousands of lines of awk that is living inside a series of giant bash scripts.
    I've seen worse, with more layers of WTF thrown in just for good measure (usually in a combination of a foolish insistence on keeping everything in one file, and a large dose of cargo cult programming — because knowing what you're doing is just so passé).

    And all wrapped in PHP. Just Because.

  • (cs) in reply to o11c
    o11c:
    Now that the contest is over, I'll link to mine (not putting it inline, that would kill the syntax highlighting) https://gist.github.com/o11c/5667876

    I would've submitted except for the "have to download and run this product" thing.

    Reminds me of infinite monkey theorem, only applied to each answer.

    ...And it probably fails if two answers start with the same character.

  • Herr Otto Flick (unregistered)

    TRWTF here is that you think we are interested in these boring articles of people trying to write stupid code. It's not funny if you try to make shit programs, it's funny when you try not to, and still manage to fuck it up.

  • cyborg (unregistered) in reply to Herr Otto Flick
    Herr Otto Flick:
    TRWTF here is that you think we are interested in these boring articles of people trying to write stupid code. It's not funny if you try to make shit programs, it's funny when you try not to, and still manage to fuck it up.

    TRWTF is haters gonna hate.

    Captcha - damnum to hell.

  • o11c (unregistered) in reply to Medinoc
    Medinoc:
    Reminds me of infinite monkey theorem, only applied to each answer.

    ...And it probably fails if two answers start with the same character.

    It works for "yes no" and "true false". Closed as not reproducible.

    On a serious note, being bug-free was not a goal of this contest. A subtler bug is that different answers do not have the same probability of being chosen

  • Jo (unregistered) in reply to o11c
    o11c:
    https://gist.github.com/o11c/5667876.

    Use /dev/random for an other WTF. Nothing like "inexplicable" hangs once in a while.

    (Captcha: similis - alike. Fitting.)

  • Slapout (unregistered)

    Where are the dishonorable mentions?

Leave a comment on “OMGWTF2: The Honorable Mentions”

Log In or post as a guest

Replying to comment #:

Log Out

« Return to Article