My RSS reader doubts MI6



  • It seems that my RSS reader does not beleave claims made by MI6




  • Did you forget to post a screenshot or something?


  • Discourse touched me in a no-no place

    @NSCoder said:

    Did you forget to post a screenshot or something?

    No they didn't. It appears your browser forgot to display the screenshot.


  • Haha, your RSS reader displays an article's title! That's funny! .....

    ... wait, I don't get it.



  • @NSCoder said:

    Did you forget to post a screenshot or something?

    [URL=http://img408.imageshack.us/my.php?image=mi6doubt2tz8.png][IMG]http://img408.imageshack.us/img408/1024/mi6doubt2tz8.th.png[/IMG][/URL]

    There should be an image above, which is a thumb linking to an image of the image as shown in my original post...

    Or go here:  http://img408.imageshack.us/my.php?image=mi6doubt2tz8.png



  • @rbowes said:

    Haha, your RSS reader displays an article's title! That's funny! .....

    ... wait, I don't get it.

     

    What I read in that image is the following conversation... 

    MI6: We did not keep files on Diana

    WTF: Do you sure?

    WTF: Not valid enough! 

     

    Or am I just odd? 



  • @GettinSadda said:

    What I read in that image is the following conversation... 

    MI6: We did not keep files on Diana

    WTF: Do you sure?

    WTF: Not valid enough! 

     

    Or am I just odd? 

     

    It's a bit of a stretch... But I got it.



  • @PJH said:

    @NSCoder said:

    Did you forget to post a screenshot or something?

    No they didn't. It appears your browser forgot to display the screenshot.

    Actually, this is a problem I noticed a while ago. The HTML for this image is:

    [code]

    [/code]

    Notice the empty <font face="Courier New">height</font> and <font face="Courier New">width</font> attributes. I use Maxthon over IE 7, and IE's engine apparently fills in zeroes, so the image shrinks to 0x0 and thus becomes invisible. I tested this in Firefox 2, and it ignored the empty attributes, showing the image. I'm not an expert in HTML, but I gathered that <font face="Courier New">height</font> and <font face="Courier New">width</font> are of the <font face="Courier New">length</font> type, and thus can't be empty, so every browser can behave as it wishes.

    Now we should determine who borked it. [b]GettingSadda[/b], what BBCodez did you use?



  • I wonder if it is the TinyMCE editor borking it?  Lemme test...

    Edit: yep, TinyMCE screws it up.  If you click the "Insert Image" button and only fill in the URL, it defaults all of the attributes.  If you use a bbcode IMG tag, it works.

    [img]http://img85.imageshack.us/img85/3223/meetingrequestgx3.png[/img]



  • @Spectre said:

    <font face="Courier New">height</font> and <font face="Courier New">width</font> are of the <font face="Courier New">length</font> type, and thus can't be empty, so every browser can behave as it wishes.
    I think it's pretty clear that the implementation most likely to render the page as the author intended will always treat an empty attribute as if it had not been specified at all.



  • @Spectre said:

    Now we should determine who borked it. GettingSadda, what BBCodez did you use?
     

    Yup, I just pressed the 'image' button and pasted in the URL - who would have thought it, a bug in the editor! 



  • @merreborn said:

    I think it's pretty clear that the implementation most likely to render the page as the author intended will always treat an empty attribute as if it had not been specified at all.

    Well, they never ask us, do they?  :'(

    Looks like this bug might be fixed in TinyMCE 3.0.

     



  • @GalacticCowboy said:

    I wonder if it is the TinyMCE editor borking it?  Lemme test...

    Edit: yep, TinyMCE screws it up.  If you click the "Insert Image" button and only fill in the URL, it defaults all of the attributes.  If you use a bbcode IMG tag, it works.

     

    Planning on taking a long Thanksgiving Holiday in 2090?

     



  • Actually, I'm still trapped in that meeting...  :(  It's been going for more than 160 days, and only 82 more years to go...  Curse those unterminated meeting requests!



  • @merreborn said:

    @Spectre said:

    <FONT face="Courier New">height</FONT> and <FONT face="Courier New">width</FONT> are of the <FONT face="Courier New">length</FONT> type, and thus can't be empty, so every browser can behave as it wishes.
    I think it's pretty clear that the implementation most likely to render the page as the author intended will always treat an empty attribute as if it had not been specified at all.

    If it was pretty clear, then W3C probably should have put that it in the standard. I don't even see how it can even be called a standard if it just says "eh, do whatever you want" in several places.


  • Discourse touched me in a no-no place

    @Cap'n Steve said:

    I don't even see how it can even be called a standard if it just says "eh, do whatever you want" in several places.
    You clearly haven't read the C or C++ ones then.



  • @Cap'n Steve said:

    If it was pretty clear, then W3C probably should have put that it in the standard. I don't even see how it can even be called a standard if it just says "eh, do whatever you want" in several places.

     

    It's called undefined behavior. Most standards have it, including almost all document, markup and programming languages. In most cases it will cause a runtime or compilation error. In error-tolerant scenarios, like rendering HTML, it leaves the application free to do what it thinks is best. In this case, IE and FF differ on what this is.



  • @mallard said:

    @Cap'n Steve said:

    If it was pretty clear, then W3C probably should have put that it in the standard. I don't even see how it can even be called a standard if it just says "eh, do whatever you want" in several places.

     

    It's called undefined behavior. Most standards have it, including almost all document, markup and programming languages. In most cases it will cause a runtime or compilation error. In error-tolerant scenarios, like rendering HTML, it leaves the application free to do what it thinks is best. In this case, IE and FF differ on what this is.

    I realize what it is, but that pretty much renders the idea of a standard pointless, doesn't it? A standard means "everyone should do things this way." If "this way" is "any way you feel like," then what's the point?



  • @Cap'n Steve said:

    @merreborn said:

    @Spectre said:

    <font face="Courier New">height</font> and <font face="Courier New">width</font> are of the <font face="Courier New">length</font> type, and thus can't be empty, so every browser can behave as it wishes.
    I think it's pretty clear that the implementation most likely to render the page as the author intended will always treat an empty attribute as if it had not been specified at all.

    If it was pretty clear, then W3C probably should have put that it in the standard. I don't even see how it can even be called a standard if it just says "eh, do whatever you want" in several places.

    Obviously you don't have much experience with W3C. 



  • @Cap'n Steve said:

    I realize what it is, but that pretty much renders the idea of a standard pointless, doesn't it? A standard means "everyone should do things this way." If "this way" is "any way you feel like," then what's the point?
     

    So you think a standard means that the creators of the standard should know every possible thing that could happen under every possible set of circumstances, and define what should happen for each and every one of those combinations of things and circumstances? Can you point me to a single existing "standard" that does this for anything non-trivial (e.g., The Standard for Collecting the Numbers One Through Nine, which requires that only the numbers 1, 2, 3, 4, 5, 6, 7, 8, and 9 are acceptable input, and that any other input is rejected, and that the input must be made using a US 101-key keyboard, a Dell Inspiron 1970 running Windows XP SP2, into an empty window in Notepad.exe. Oops! That user used their left hand to type the digit '1' on the numeric keypad on the RIGHT side of the keyboard. Is that allowed? No? What should we do in that case? Is the numeric keypad even allowable? Did they install all of the possible combinations of files when the SP2 update was done?)

    A standard defines certain conditions that can happen, and the behavior expected when that condition occurs. It's impossible to think of every possibility and then define behavior for those possibilities; the standard would be infinitely under development (and therefore not a standard). You can't stop everything just because the standard doesn't define what you should do when stuff happens that they didn't anticipate during development of the standard. Undefined behavior has to be expected and accepted; the standard can be updated to a newer version later.



  • @Cap'n Steve said:

    @mallard said:

    @Cap'n Steve said:

    If it was pretty clear, then W3C probably should have put that it in the standard. I don't even see how it can even be called a standard if it just says "eh, do whatever you want" in several places.

     

    It's called undefined behavior. Most standards have it, including almost all document, markup and programming languages. In most cases it will cause a runtime or compilation error. In error-tolerant scenarios, like rendering HTML, it leaves the application free to do what it thinks is best. In this case, IE and FF differ on what this is.

    I realize what it is, but that pretty much renders the idea of a standard pointless, doesn't it? A standard means "everyone should do things this way." If "this way" is "any way you feel like," then what's the point?

    If a standard specified [i]everything[/i], it would be a) orders of magnitude thicker, b) mind-numbingly boring, and, most significantly, c) no one will implement it, because it will defy creativity and/or common sense. And besides, it it supposed that everybody will make proper HTML ([b]cough[/b]), so the undefined behavior will be avoided.



  • @Cap'n Steve said:

    I realize what it is, but that pretty much renders the idea of a standard pointless, doesn't it? A standard means "everyone should do things this way."
     

    W3C standards are more of a "best practices" thing anyways, since their vetting process works on stone-age time, and browsers in the 90s were developing at internet speeds. W3C couldn't keep up, so the html 3 & 4 specs were "here's the average of what everyone's doing". It wasn't until XML and XHTML start coming out that they became somewhat proactive, and even there they're way behind the times. 



  • @KenW said:

    TL;DR

    I suggest you could have made your point better by simply questioning the good of a standard defining the correct behaviour of empty dimension properties ...



  • @KenW said:

    A standard defines certain conditions that can happen, and the behavior expected when that condition occurs. It's impossible to think of every possibility and then define behavior for those possibilities; the standard would be infinitely under development (and therefore not a standard). You can't stop everything just because the standard doesn't define what you should do when stuff happens that they didn't anticipate during development of the standard. Undefined behavior has to be expected and accepted; the standard can be updated to a newer version later.

    Right, but in this case (and I know I've seen other examples of this in HTML) they did think of the possibility; they just decided for some reason not to specify what should happen.


Log in to reply