• accident (unregistered) in reply to Zylon

    if(rand() % 10 < 2) comment.makeFeatured();

  • What (unregistered) in reply to Outlaw Programmer

    Wow never met someone that has worked at every company in the world! You must be a busy man.

    I have worked at several companies as a lone developer where they have insisted that I spend extra time to fix the problem at hand as opposed to hacking a work around.

  • Randomosity (unregistered) in reply to Jimmy
    Jimmy:
    66.77.93.50 doesn't resolve to a DNS entry. What's really going on here?

    OHNOES! TEH INTARNETZ R DOWN!

    Or: The names and IP addresses have been changed to protect the clueless.

  • (cs) in reply to mathew
    mathew:
    It's really very simple.

    The browser requests the resource at foo.com/favicon.ico. The server responds with a 302 response, which means "That thing you asked for is temporarily over here -> /error.aspx". The new URL is for a custom error page, but THE BROWSER DOESN'T KNOW THAT, because the server hasn't told it there's an error. It has just told it that the resource has moved. So the browser requests foo.com/error.aspx as the URL of the favicon instead, like it has been told. The server opens a new session and fails, and issues another 302 redirect saying "That thing you asked for is temporarily over here -> /access?action=forward&uri=%2Ferror.aspx" Again, the browser doesn't know that's an error page, so it dutifully follows the instructions the server has given it. And by so doing, another session is created, and the infinite loop begins.

    The only fault of the browser is that arguably, it should limit the number of successive 302 redirects it will follow for a given resource being requested, and eventually give up. However, the standard (RFC1945) doesn't say anything about such a limit being required, or even desirable.

    So although everybody loves to hate on AOL, in this case the fault is absolutely entirely with the server and application.

    I've seen similar things myself. I've browsed to web sites that try to start a session, then fail because I have cookies disabled by default, so they redirect me back to the page to start a session, and so on forever. I tend to leave pages like that open in another tab for a few hours in the hope that the fucktards running the site will notice the problem.

    Of course the real solution is to return the custom error page in the original response with a 404 code, instead of redirecting to the error page. This lets the browser know it's an error, so if it's not interested in errors it just closes the connection (it doesn't even need to wait for the entire response), instead of dutifully following the redirect.

    Of course, having it be possible for the server to enter an infinite redirect loop is more of a wtf than redirecting to an error page.

  • Steve H. (unregistered)

    Let's see if I can point out all the WTFs:

    1. AOL's browser opens a new session for each HTTP redirects 302 error and fails to close the previous session.

    2. The company in question requires the use of cookies to properly generate the appropriate error message, though I'm not sure how to even set something like that up.

    3. Same said company doesn't utilize any form of session limiting.

    The real funny part is that there's a random family in Ohio who's browser response time probably got sped up by Bob B.

  • (cs) in reply to DeLos

    Ladies and gentlemen, allow me to present a product of public education.

    DeLos:
    What is an "Ohio"?
  • iToad (unregistered)

    If your car was designed like some websites, it would explode into a huge fireball of death every time that a taillight burned out.

  • (cs) in reply to ssprencel

    Guess this is one of those days where EVERYONE'S sarcasm detector is broken.

    I didn't mean to imply that this guy shouldn't have provided a real fix for the bug, or that his management is incompetent. In fact, it's quite the obvious in this case because even though they were afraid of introducing new bugs, they authorized this guy to find and fix the problem.

    My point was that, sometimes, there are bugs in the system that are really annoying but never get fixed because there is a workaround. Some managers think "why should we fix this bug when it's not really a problem? We can't sell a bugfix, so let's just keep working on new features!"

    Another point is that sometimes people try to fix things with a Rube Goldberg-ian solution that just complicates matters. "Hey, our workaround is to just restart these services, so let's write a program/script that does that for us and hope that THIS program doesn't have any bugs either!"

    Geez, some of you guys need some perspective here. The great thing about this site is that most of the WTFs can be generalized and applies to everyone's work place. I didn't ask Alex to make my comment a features one. I can only guess that he saw the point that I was trying to make and got a chuckle out of it.

  • darryl (unregistered) in reply to DeLos

    It wasn't one person, AOL at one time had all their users coming out of a point on Ohio.

  • 008 (unregistered) in reply to elias
    elias:
    Botzinger Gulm:
    CynicalTyler:
    Jason:
    Mattkins:
    There. Fixed that for you.
    There. Fixed that for you.
    There. Fixed that for you.
    There. Fixed that for you.
    There. Fixed that for you.
    Exception in thread "main": StackOverflowError at WTF.generateUselessQuote (Native Method) at WTF.generateUselessQuote (Native Method) at WTF.generateUselessQuote (Native Method) ...
  • Christopher (unregistered) in reply to Seraph
    Seraph:
    Outlaw Programmer:
    I call shenanigans. Management would have just told the developer to schedule a task that restarts the 2 services every night. At the places where I've worked, any bug that has a workaround never gets fixed; the workaround just gets added to the user manual!
    From the description it doesn't sound like that would have fixed the bug.

    [Joke]

    O <--- Your head.

  • (cs) in reply to Outlaw Programmer
    Outlaw Programmer:
    I didn't mean to imply that this guy shouldn't have provided a real fix for the bug, or that his management is incompetent. In fact, it's quite the obvious in this case because even though they were afraid of introducing new bugs, they authorized this guy to find and fix the problem.

    My point was that, sometimes, there are bugs in the system that are really annoying but never get fixed because there is a workaround. Some managers think "why should we fix this bug when it's not really a problem? We can't sell a bugfix, so let's just keep working on new features!"

    Another point is that sometimes people try to fix things with a Rube Goldberg-ian solution that just complicates matters. "Hey, our workaround is to just restart these services, so let's write a program/script that does that for us and hope that THIS program doesn't have any bugs either!"

    You are assuming that he asked for permission before fixing it. The article only said he got permission to look into it as long as he wasn't too intrusive.

  • (cs) in reply to ssprencel

    Ladies and gentlemen, allow me to present a product of public education.

    ssprencel:
    Ladies and gentlemen, allow me to present a product of public education.
    Fixed that for you.
  • Tei (unregistered) in reply to Jimmy
    Jimmy:
    66.77.93.50 doesn't resolve to a DNS entry. What's really going on here?

    No idea!.

    Call 555-9030200300112

  • (cs)

    The Real WTF is that the browser followed the redirects into infinity. Doesn't the HTTP spec recommend a limit of five redirects after which the browser should call it quits?

    Then again, AOL runs with IE under the hood...

  • Tei (unregistered) in reply to PSWorx
    PSWorx:
    The Real WTF is that the browser followed the redirects into infinity. Doesn't the HTTP spec recommend a limit of five redirects after which the browser should call it quits?

    Then again, AOL runs with IE under the hood...

    I quote your message, but I will not make a reply.

    I used to download stuff from internet on the university at night. This mean a huge pipe, almost no users.

    So a recursive wget command made sense.

    like

    wget -r -l 32 http://www.geocities.com/something/something...

    Thats whas the night I mirrored geocities.

    Other night I tried to download a web search engine with a recursive link. Thats how you create a directory that mighnight commander can't delete because has too much recursive folders inside folders inside folders.. in a ex3 (or whas reiser?) partition.

    University, oh.. good times :D

  • (cs) in reply to Outlaw Programmer
    Outlaw Programmer:
    Guess this is one of those days where EVERYONE'S sarcasm detector is broken.

    I didn't mean to imply that this guy shouldn't have provided a real fix for the bug, or that his management is incompetent. In fact, it's quite the obvious in this case because even though they were afraid of introducing new bugs, they authorized this guy to find and fix the problem.

    My point was that, sometimes, there are bugs in the system that are really annoying but never get fixed because there is a workaround. Some managers think "why should we fix this bug when it's not really a problem? We can't sell a bugfix, so let's just keep working on new features!"

    Another point is that sometimes people try to fix things with a Rube Goldberg-ian solution that just complicates matters. "Hey, our workaround is to just restart these services, so let's write a program/script that does that for us and hope that THIS program doesn't have any bugs either!"

    Geez, some of you guys need some perspective here. The great thing about this site is that most of the WTFs can be generalized and applies to everyone's work place. I didn't ask Alex to make my comment a features one. I can only guess that he saw the point that I was trying to make and got a chuckle out of it.

    For the record, I liked your comment and think it deserves to be featured.

  • He Who Is Looking For Trouble (unregistered) in reply to Outlaw Programmer

    You must work for IBM too

  • apaq11 (unregistered) in reply to Outlaw Programmer
    Outlaw Programmer:
    I call shenanigans. Management would have just told the developer to schedule a task that restarts the 2 services every night. At the places where I've worked, any bug that has a workaround never gets fixed; the workaround just gets added to the user manual!

    You must work in every software company ever...

  • Troy McClure (unregistered) in reply to apaq11
    begin
      if comment_contains_trwtf then
        featured := false;
      elsif comment_mentions_goggles then
        featured := false;
      elsif comment_mentions_wooden_table then
        featured := false;
      elsif post_mentions_captcha then
        featured := false;
      else
        featured := true;
      end if;
    end;
    
  • Watson (unregistered) in reply to iToad
    iToad:
    If your car was designed like some websites, it would explode into a huge fireball of death every time that a taillight burned out.
    That certainly is the sort of design thinking that lies behind starting a whole entire user session just because the favicon was being requested.
  • Robin Goodfellow (unregistered) in reply to Tei
    Tei:
    I agree. And thats my problem with the current WTF, the solution is another WTF!, the problem is just rub under the carpet with a touch favicon.ico, but still here, waiting to happend again.

    Read closer, the solution wasn't to add the favicon.ico, the solution was to fix the error redirect loop. The addition of the favicon.ico was just a "while we're at it, might as well avoid a bunch of 404s in the logs in the future" sort of thing.

  • (cs) in reply to ssprencel
    ssprencel:
    Ladies and gentlemen, allow me to present a product of public education.
    DeLos:
    What is an "Ohio"?

    Or...a non-American?

  • (cs) in reply to Someone You Know
    Someone You Know:
    Or...a non-American?
    Or someone who didn't count on all the broken sarcasm detectors.
  • (cs)

    Hey, ho, way to go Ohio!

  • Say No to Cowboy Programming (unregistered) in reply to Outlaw Programmer

    You are right for many cowboy programming environments but not right for shops where quality matters. I've been in both shops and the cowboy programming shops let the problem go. The Agile, Iterative or disciplined shops do not allow crap like this to linger, let alone go on with a ridiculous loser solution like auto restart IIS. That is for kindergartners.

  • (cs) in reply to Jimmy
    Jimmy:
    66.77.93.50 doesn't resolve to a DNS entry. What's really going on here?

    Anonymization, holmes

  • notme (unregistered) in reply to Anony-mouse
    Anony-mouse:
    Ah ASP. PHP or mod_rewrite would automatically kill these loops after a set number of redirects.

    Really? That's news to me, and I have worked with both of these things.

    Okay, it might work with transparent redirects (where the URL change visible to the user does not change) with mod_rewrite, but not with anything that involves a "Location: " in the HTTP-header or a <meta http-equiv="refresh" ...>, unless you manually carry a counter of some sort in the URL.

    Also, I have no idea what difference there would be between ASP and PHP in this regard.

  • notme (unregistered) in reply to notme
    notme:
    Anony-mouse:
    Ah ASP. PHP or mod_rewrite would automatically kill these loops after a set number of redirects.

    Really? That's news to me, and I have worked with both of these things.

    Okay, it might work with transparent redirects (where the URL change visible to the user does not change) with mod_rewrite, but not with anything that involves a "Location: " in the HTTP-header or a <meta http-equiv="refresh" ...>, unless you manually carry a counter of some sort in the URL.

    Also, I have no idea what difference there would be between ASP and PHP in this regard.

    :s/URL change visible/URL visible/g

  • Zock (unregistered) in reply to Jimmy
    Jimmy:
    66.77.93.50 doesn't resolve to a DNS entry. What's really going on here?

    Not only that, but it's more likely to be somewhere in California. (I knew there wasn't any Internet in Ohio!)

  • (cs) in reply to Tei
    Tei:
    Jimmy:
    66.77.93.50 doesn't resolve to a DNS entry. What's really going on here?

    No idea!.

    Call 555-9030200300112

    No, call 0-118-9998819991197253!

  • (cs) in reply to Jimmy
    Jimmy:
    66.77.93.50 doesn't resolve to a DNS entry. What's really going on here?
    Sarcasm?

    Not every IP address has a PTR entry (and not every IP address warrants a PTR entry if it resolves to more than one host; think NAT). The address above may have been changed to protect the innocent (guilty?) as it belongs to a half "Class-C" block (it falls in the range of Class-B addresses) owned by 101 Communications (http://www.101com.com) based in Chatsworth, CA.

    Qwest Communications Corporation QWEST-INET-12 (NET-66-77-0-0-1) 
                                      66.77.0.0 - 66.77.255.255
    101 Communications LLC QWEST-BUC-101COMM1 (NET-66-77-93-0-1) 
                                      66.77.93.0 - 66.77.93.127
    

    I'd say it's nothing more than editorial license.

  • David Schwartz (unregistered)

    Something doesn't seem right. If the problem with the favicon file was simply that it didn't exist, the problem would reappear as soon as someone else tried to retrieve a file that didn't exist.

    It's hard to imagine that doesn't happen all that often. Even the very smallest websites see the occasional "URL from space" in their logs.

  • (cs) in reply to ssprencel
    ssprencel:
    Ladies and gentlemen, allow me to present a product of public education.
    DeLos:
    What is an "Ohio"?

    The real WTF is that some people assume that everyone in the world knows everything about the USA

  • David Schwartz (unregistered) in reply to David Schwartz

    I ignore me. I see that adding the favicon file wasn't his fix for the problem.

  • PJ (unregistered) in reply to James
    James:
    DeLos:
    sweavo:
    Wait... when did Ohio get internet?

    What is an "Ohio"?

    I believe it means "Hello" in Japanese.

    Technically, means "Good Morning" if you're going for a literal kinda thing.

    Which, I believe, makes it the one and only appropriate use of the word "good" in context with the word "Ohio".

  • (cs) in reply to ssprencel
    ssprencel:
    Ladies and gentlemen, allow me to present a product of public education.
    DeLos:
    What is an "Ohio"?

    You guys aren't kidding about sarcasm detectors being off.

    If you read ALL of the comments (in order) then maybe you follow along.

  • (cs)

    And yes, I was taught in public schools. One of the better ones in my home state (not Ohio). Private College for my real book learning though.

    If that is ok...

  • (cs) in reply to Tei
    Tei:
    I quote your message, but I will not make a reply.
    Do you want to tell me something?
  • mathew (unregistered) in reply to Thief^
    Thief^:
    Of course, having it be possible for the server to enter an infinite redirect loop is more of a wtf than redirecting to an error page.

    Want to see a major web site set up to go into an infinite loop right now?

    Turn off cookies in Firefox and go to mysimon.com.

    (Or have Firefox set to ask, and say no, you don't want to accept cookies from that site.)

    Firefox and MySimon have been quietly infinite looping for about 5 minutes now.

  • (cs)

    To be fair, even DTWTF itself had the occasional infinite loop in the past. One incident was especially intriguing, where each iteration added another "forums." subdomain to the URL, resulting in something like "forums.forums.forums.forums.forums.forums.forums.thedailywtf.com" rather soon. Needless to say, hilarity ensued...

  • (cs) in reply to Vombatus
    Vombatus:
    ssprencel:
    Ladies and gentlemen, allow me to present a product of public education.
    DeLos:
    What is an "Ohio"?

    The real WTF is that some people assume that everyone in the world knows everything about the USA

    The real WTF is that (according to rumor) most people in the world know more about the USA (geographically) than most US citizens. And let's not even talk about the rest of the world.

  • Greg (unregistered)

    I'm of the opinion that while foisting all the blame on this one family from Ohio, the programmer in this case should have perhaps been examining the code that allowed 2 million sessions with the DB to open from one client. You're running an e-commerce site which gets 1,000 visitors per day, the SQL server is running out of resources for new SQL sessions and your solution is to simply allocate more resources? Even the most cursory investigation should have led him to realise that a very simple (and nowadays usually inbuilt) optimisation technique like connection pooling would significantly reduce the number of sessions to users ratio. Its maybe even possible that connection pooling was in use and he was forgetting to close the database connection in his code. There's also the question of how the hell did a request for a non-existent resource lead to an infinite redirect loop? There's no way you can blame that one on IIS or the user, that's a fault in the code. Really, one user should not have been able to bring the server to its knees like that; I say the blame lies with the server admins and programmers here.

  • John Doe (unregistered) in reply to Randomosity
    Randomosity:
    Jimmy:
    66.77.93.50 doesn't resolve to a DNS entry. What's really going on here?

    OHNOES! TEH INTARNETZ R DOWN!

    Or: The names and IP addresses have been changed to protect the clueless.

    In this case apparently anonymization of an Iranian family, because it can't be resolved due the cable cuts.

  • James Mayfield AU (unregistered) in reply to Outlaw Programmer
    Outlaw Programmer:
    the workaround just gets added to the user manual!

    I call shenanigans on your shenanigans.

    At the places where I've worked, any bug that has a workaround never gets the workaround added to the user manual. The updating just gets added to someones low-priority task list and never sees the light of day.

  • (cs) in reply to Greg
    Greg:
    I say the blame lies with the server admins and programmers here.
    Congratulations on saying what everybody else has already been saying.
  • Jonathan (unregistered) in reply to DeLos
    DeLos:
    sweavo:
    Wait... when did Ohio get internet?

    What is an "Ohio"?

    An "Ohio" would be an Ohio class ballistic missile submarine.

    And I didn't know SSBNs had gotten Internet access yet.

  • (cs)

    These people aren't real programmers, are they?

  • (cs) in reply to n9ds
    n9ds:
    Hey, ho, way to go Ohio!
    As in, "Four dead in..."?

    No, sorry, I was just pretending. How're the parking lots there nowadays?

  • GB (unregistered)

    At the risk of sounding clueless: the lack of support for cookies was enough for the server to have a new session created on each request from that client?

Leave a comment on “The Most Favoritest Icon”

Log In or post as a guest

Replying to comment #:

« Return to Article