• Sarah (unregistered)

    Sorry I was distracted by the cute kitten picture, what the article about again?

  • NaN (unregistered)

    And look at Hebrew University of Jerusalem coming in at #1 in Publications category (when reading right to left anyway)

  • (cs)

    Toastmasters' central organising body don't really get the Web. Just last year they spent a considerable amount of money on a new reporting system in Silverlight, so I wouldn't be surprised if the website is only tested in IE.

    And whichever ranking organisation that is doesn't get the concept of ranks. There are 4 organisations scoring 100 for "Education", and they rank them 1, 2, 3, and 4. Joint first is clearly too complex.

  • (cs)

    I'd say that the Windows Azure form Works as advertised. Maybe the warning should be placed below the comment field, but clearly someone input 1 character (look at Count below field) and the validator needs at least 5. Maybe the shortest conceivable sentence is "Fuck!"?

  • qdfqdfqsdfqsdfsdf (unregistered) in reply to dsckeld
    dsckeld:
    I'd say that the Windows Azure form Works as advertised. Maybe the warning should be placed below the comment field, but clearly someone input 1 character (look at Count below field) and the validator needs at least 5. Maybe the shortest conceivable sentence is "Fuck!"?

    What count below field? Are you refering to '7999 of 8000 characters'?

    Note that it does not say: '7999 of 8000 characters remaining'.

    Looking at the screenshot should make it very clear that both fields contain more then 5 characters. (some of the words in the screenshot are readable)

  • Warren (unregistered)

    At least we know why Jeff Bezos chose the Washington Post over TechCrunch.

  • (cs)
    Alex N.:
    I was thrilled to see we here in Oregon rank favorably, at the very top of the top 26 list
    At the very top, yes. But you are not number 1.

    You are zero with zero publications, zero citations, zero publications, zero education.

    Must suck to be employed there - wait, there are also no employees!

  • Bob (unregistered)

    Why Jeff Bezos Bought The Washington Post: Awwwwwwww!

  • The Mole (unregistered)

    The list is also demonstrating an American's concept of geography. Whilst England is technically a country (the best type of right) I think Oxford and Cambridge would prefer being ranked as being the best Universities in the whole of the UK not just England.

  • omg (unregistered)

    is that...

    is that Android KitKat? :O

  • QJo (unregistered)

    Obviously a typo for Espa 'n' Tilde, the well-known comedy/singing duo.

  • (cs) in reply to Bob
    Bob:
    Why Jeff Bezos Bought The Washington Post: Awwwwwwww!
    Yep, clearly he bought it to introduce more photos of kittens.
  • Chris Morgan (unregistered)

    The form submission maximum length issue is one where I believe I figured out the answer a year or two back when I came across a similar disparity between stated and actual limit and decided I must work it out.

    The answer is line breaks: the browser treats a line break as a single character (CR, \n) before you submit, but submits it as two characters (CR LF, \r\n).

    Don't believe me? Here's the proof:

    <a rel="nofollow" href="data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>" target="_blank" title="data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>">data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>

    Thus, if you include a couple of line breaks, your 7999 characters is actually sent as 8001 characters.

  • Chris Morgan (unregistered) in reply to Chris Morgan
    Chris Morgan:
    The answer is line breaks: the browser treats a line break as a single character (CR, \n) before you submit, but submits it as two characters (CR LF, \r\n).

    Whoops, I meant LF for \n, not CR.

  • (cs)

    So does this mean that pictures of kittens has now surpassed porn as the reason for the internet? Or does this mean that now that kittens are about to take over, they will merge with porn and take over forever. Oh Noes! Kitty Pron!

  • anonymous (unregistered) in reply to Chris Morgan
    Chris Morgan:
    The form submission maximum length issue is one where I believe I figured out the answer a year or two back when I came across a similar disparity between stated and actual limit and decided I must work it out.

    The answer is line breaks: the browser treats a line break as a single character (CR, \n) before you submit, but submits it as two characters (CR LF, \r\n).

    Don't believe me? Here's the proof:

    <a rel="nofollow" href="data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>" target="_blank" title="data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>">data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>

    Thus, if you include a couple of line breaks, your 7999 characters is actually sent as 8001 characters.

    Yeah, but asking him to submit that as a new problem using the same form is hardly a WTF. If he couldn't express that problem ("Trying to submit problems of slightly under 8000 characters fails and says the limit is 5-8000 characters") in less than 8000 characters (minus however many line breaks he used), I think that's probably TRWTF.

  • ac (unregistered) in reply to Chris Morgan

    Also encoding differences can do this, but very similar issue, 1 character from js view becomes 2 or more to the back end.

  • (cs) in reply to The Mole
    The Mole:
    The list is also demonstrating an American's concept of geography. Whilst England is technically a country (the best type of right) I think Oxford and Cambridge would prefer being ranked as being the best Universities in the whole of the UK not just England.

    If it had said UK, there would have been pedantic Brits complaining about that, as well.

  • (cs) in reply to The Mole
    The Mole:
    The list is also demonstrating an American's concept of geography. Whilst England is technically a country (the best type of right) I think Oxford and Cambridge would prefer being ranked as being the best Universities in the whole of the UK not just England.
    Yeah... because if someone doesn't know Oregon is in the USA, they must be an American.

    You're a troll.

  • Dave (unregistered) in reply to no laughing matter
    no laughing matter:
    Alex N.:
    I was thrilled to see we here in Oregon rank favorably, at the very top of the top 26 list
    At the very top, yes. But you are not number 1.

    You are zero with zero publications, zero citations, zero publications, zero education.

    Must suck to be employed there - wait, there are also no employees!

    They're ranked zero because of their student employee relations policy.

  • ¯\(°_o)/¯ I DUNNO LOL (unregistered)

    We're number zero! We're number zero! In your face, space coyote!

  • (cs) in reply to pjt33
    pjt33:

    And whichever ranking organisation that is doesn't get the concept of ranks. There are 4 organisations scoring 100 for "Education", and they rank them 1, 2, 3, and 4. Joint first is clearly too complex.

    Roundoff error, duh.

  • Meep (unregistered) in reply to pjt33
    pjt33:
    Toastmasters' central organising body don't really get the Web. Just last year they spent a considerable amount of money on a new reporting system in Silverlight, so I wouldn't be surprised if the website is only tested in IE.

    They've had larger concerns than the inter-tubes, seeing as how they've been a little busy mastering toast.

  • (cs) in reply to Chris Morgan
    Chris Morgan:
    Chris Morgan:
    The answer is line breaks: the browser treats a line break as a single character (CR, \n) before you submit, but submits it as two characters (CR LF, \r\n).

    Whoops, I meant LF for \n, not CR.

    Well actually '\n' is NEW LINE, but it is encoded as <LF> because the character is actually NEW LINE/LINE FEED.

    If you go to EBCDIC, it is 0x15, but nobody really cares about EBCDIC anymore.

  • (cs) in reply to Bob
    Bob:
    Why Jeff Bezos Bought The Washington Post: Awwwwwwww!

    "I think it would be fun to run a newspaper."

  • (cs)

    With a little better table design, the top 26 list could have shown "OREGON INSTITUTE OF TECHNOLOGY / OREGON / null / null / null / null / null / null / null / null / null / null / null / null / null". In DB2, this would have sorted dead last.

    Which raises a question: Is it better to be a big 'zero' at the top or a big 'null' at the bottom?

  • Jellineck (unregistered)

    "Nothing works in this browser"

    You could just download another browser. No reason to 86 the whole phone.

  • JocularJuggler (unregistered) in reply to The Mole

    Re: Internet is Broken - Time to Get a New Phone 2013-10-25 07:29 • by The Mole The list is also demonstrating an American's concept of geography. Whilst England is technically a country (the best type of right) I think Oxford and Cambridge would prefer being ranked as being the best Universities in the whole of the UK not just England.

    Or at least the best in Great Britain.

  • (cs)

    I'm actually in "Espa'&'ntilde;ol" class right now. I'm supposed to be working on some assignments on the computer.

  • Nagesh (unregistered) in reply to Warren
    Warren:
    At least we know why Jeff Bezos chose the Washington Post over TechCrunch.
    And still, that was the most informative and enjoyable TechCrunch article I have ever read.
  • n_slash_a (unregistered)

    TRWTF is how Harvard is still ranked 100 out of 100 in almost all of the categories.

  • (cs) in reply to ac
    ac:
    Also encoding differences can do this, but very similar issue, 1 character from js view becomes 2 or more to the back end.
    But here at the top-ranked Oregon Institute of Technology, Español becomes Espa'&'ntilde;ol, an increase of nine characters.

    And that's why Jeff Bezos bought the Washington Post.

  • El-Ka-Ben (unregistered) in reply to Chris Morgan
    Chris Morgan:
    The form submission maximum length issue is one where I believe I figured out the answer a year or two back when I came across a similar disparity between stated and actual limit and decided I must work it out.

    The answer is line breaks: the browser treats a line break as a single character (CR, \n) before you submit, but submits it as two characters (CR LF, \r\n).

    Don't believe me? Here's the proof:

    <a rel="nofollow" href="data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>" target="_blank" title="data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>">data:text/html,<form action=http://httpbin.org/post method=post><textarea name=foo id=foo onkeyup='submit.value="Length: "+foo.value.length'></textarea><input type=submit id=submit></form>

    Thus, if you include a couple of line breaks, your 7999 characters is actually sent as 8001 characters.

    That's a fairly useful thing to know... I wouldn't have expected it to count line breaks at all, but those are the quirks. Some languages/functions don't count any whitespace.

    I usually try to skip whitespace when I do a count, it's generally what the end-user expects.

  • (cs) in reply to pjt33
    pjt33:
    Toastmasters' central organising body don't really get the Web. Just last year they spent a considerable amount of money on a new reporting system in Silverlight, so I wouldn't be surprised if the website is only tested in IE.
    People actually use silverlight?
  • (cs) in reply to Coyne
    Coyne:
    With a little better table design, the top 26 list could have shown "OREGON INSTITUTE OF TECHNOLOGY / OREGON / null / null / null / null / null / null / null / null / null / null / null / null / null". In DB2, this would have sorted dead last.

    Which raises a question: Is it better to be a big 'zero' at the top or a big 'null' at the bottom?

    Big null at the bottom of course, bozo, 'cos then people won't see you embarrassing yourself up at the top like a prick in a convent.
  • (cs)

    Maybe that Bezos article was posted with something from http://placekitten.com/

    Don't have a photo? They'll serve up a kitten to fit your dimensions. Yikes...I mean awwwwwww.

  • J (unregistered)

    I'm pretty sure you aren't suppose to use these pages anymore: http://www.toastmasters.org/FooterMenuCategories/BrowserCompatibility.aspx

  • harryhashtable (unregistered) in reply to Coyne
    Coyne:
    With a little better table design, the top 26 list could have shown "OREGON INSTITUTE OF TECHNOLOGY / OREGON / null / null / null / null / null / null / null / null / null / null / null / null / null". In DB2, this would have sorted dead last.

    Which raises a question: Is it better to be a big 'zero' at the top or a big 'null' at the bottom?

    NULL means unknown, whereas zero is a definitive value. That makes NULL the referred value, could be first or last, the rank is unknown. Oregon may well have scored zero.

  • (cs) in reply to Zecc
    Zecc:
    Bob:
    Why Jeff Bezos Bought The Washington Post: Awwwwwwww!
    Yep, clearly he bought it to introduce more photos of kittens.
    No. That kitten was at the center of a hidden scandal with Bozoes. He was blackmailed into buying it. Why else would anyone buy the Post?
  • P1xels (unregistered)

    Woah, that skype video message image is actually mine! http://imgur.com/J5GPyRN Never thought I'd get on here!

  • UKSCEPTIC (unregistered) in reply to JocularJuggler

    UK = GB + Northern Ireland GB = England + Scotland + Wales

  • (cs) in reply to UKSCEPTIC
    UKSCEPTIC:
    UK = GB + Northern Ireland GB = England + Scotland + Wales
    Man?
  • Ted (unregistered)

    If Georgia can be a country, why not Oregon?! :)

  • C (unregistered) in reply to pjt33
    pjt33:
    And whichever ranking organisation that is doesn't get the concept of ranks. There are 4 organisations scoring 100 for "Education", and they rank them 1, 2, 3, and 4. Joint first is clearly too complex.
    Actually, they seem to have used to overall score as tie-breaking criteria...
  • qbolec (unregistered) in reply to Chris Morgan
    Chris Morgan:
    The answer is line breaks: the browser treats a line break as a single character (CR, \n) before you submit, but submits it as two characters (CR LF, \r\n).
    It's even funnier as it varies between browsers/OSes, which causes funny subtle bugs when we try to match @mentions and #hashtags found in text to their positions (offsets inside the string) -- often JS sees that differently than the webserver, and the thing gets further complicated as obviously it works differently for regular <form> POST, and differently for AJAX XHR. We end up with <input type=hidden name="was_counted_with_crs"> with value set by javascript once JS detects if that particular browser counts CR LF in textarea as 2 or 1 bytes. It is even more funny if you try to assign a string with newlines to a textarea.value, or textarea.innerHTML or textarea.innerText (especially in old IE). Even appending a TextNode does not always work as expected.
  • qbolec (unregistered)

    Let me elaborate.

    In older IE : e.innerHTML = "a b";//two spaces e.innerHTML.length == 3 //IE treats any whitespaces as single space e.innerHTML = "a\nb"; e.innerHTML === "a b" //yup, even \n gets replaced as space

    This affects for example Mootool's new Element({html:"a b"}).

    Similar problem with e.set('text',"a\nb") and new Element({text:"a\nb"});

    A workaround is to perform e.appendChild(document.createTextNode(x.replace(/\r\n/g,'\n')); where as you see, you have to remove \r, because on some versions of IE there is a quite opposite bug: a line becomes two lines, as each of two characters (\r and \n) are treated as 2 separate newlines. I know it's crazy, but it got's crazier...

    Consider this snippet (IE again):

    document.body.innerHTML = ''; var x = new Element('span',{ 'style': 'white-space: pre' }).inject(document.body); x.set('html','a\nb')

    you may SEE that it LOOKS in the browser as two separate lines, BUT document.getElementsByTagName('span')[0].innerHTML returns a string with length 2, that is exactly "ab". Yup, the \n gets eaten entirely, when reading the property.

    Now...if you put the set inside a timeout, like this:

    document.body.innerHTML = ''; var x = new Element('span',{ 'style': 'white-space: pre' }).inject(document.body); setTimeout(function(){ x.set('html','a\nb') },0 )

    so that we defer html assignment until the current callstack exits and the injection is "totally done", we get tottaly different results. Namely, we SEE a single line in the browser, and, moreover document.getElementsByTagName('span')[0].innerHTML returns 3 letter string "a b". Yup, the newline got replaced with a space this time.

    Now... this was just about the innerHTML and innnerText properties of textarea/span/pre... similar stories about .value haunt me as well. For example, standard http://www.w3.org/TR/html4/interact/forms.html#h-17.13.4.1 w application/x-www-form-urlencoded says that
    Line breaks are represented as "CR LF" pairs (i.e., `%0D%0A') when submited to server. On the other hand inside javascript strings you can spot "\n" or "\r\n" line endings depending on your OS/Browser http://stackoverflow.com/questions/1155678/javascript-string-newline-character . It may also depend on the source of the string -- if you read it from element.value it may be quite different than if you simply created it yourself. Aggrh.

  • UKSCEPTIC (unregistered) in reply to Kivi

    Isle of Man is a self-governing British Crown Dependency.

    It's within the British Isles (as is Jersey, Gurnsey, Republic of Ireland) which is a geographical not a political area, not part of UK or GB.

  • Matthew (unregistered)

    TechCrunch is making fun of the Chicago Tribune, who made the mistake for real, on their front page. See 'techcrunch.com/2013/08/09/headline-test-here' for the Test test article, which links to 'gizmodo.com/the-chicago-tribune-has-made-the-best-internet-mistake-964073520'.

  • (cs) in reply to UKSCEPTIC
    UKSCEPTIC:
    Isle of Man is a self-governing British Crown Dependency.

    It's within the British Isles (as is Jersey, Gurnsey, Republic of Ireland) which is a geographical not a political area, not part of UK or GB.

    Oh, like this! (Not original) [image]

  • Your Name (unregistered)

    "I am error."

    Nice Zelda II reference.

Leave a comment on “Internet is Broken - Time to Get a New Phone”

Log In or post as a guest

Replying to comment #:

Log Out

« Return to Article