Re: very very sad: most browsers are broken :(

Discussion in 'HTML' started by Bruce Grubb, Apr 18, 2004.

  1. Bruce Grubb

    Bruce Grubb Guest

    In article <>,
    Gregory Weston <> wrote:

    > In article <>,
    > Bruce Grubb <> wrote:
    >
    > > In article <170420040759517445%>,
    > > Thomas Reed <> wrote:
    > >
    > > > In article <-berlin.de>, Joe Doe
    > > > <> wrote:
    > > >
    > > > > PS: The validator page itself http://validator.w3.org returns 9 errors
    > > > > about its own html code. "But I read it in the paper!..."
    > > >
    > > > A validator really gives you very little useful information except
    > > > whether your page conforms to standards. Since almost all web browsers
    > > > render non-standard pages, this is not useful information, IMHO.

    > >
    > > Actually it is VERY useful information as by going to the standard rather
    > > than some browser glitches make is more likely (though not certain) that
    > > the thing to be read correctly by a 4.01 complient browser you have not
    > > even heard of.

    >
    > A problem arises, however, if an exceedingly popular browser does not
    > render correct HTML in a worthwhile fashion.


    Now what 'exceedingly popular browser' does this? Remember HTML is NOT a
    page layout format (that is what PDF is for) so a specific layout does NOT
    count.

    Back in the "who uses virtual pc for testing webpages" thread in alt. html
    and comp.infosystems.www.authoring some months back the only examples
    people could provide fell into several categories:


    1) Browser detection crap (javascript, java, etc)

    2) Minor display hiccups (HTML is NOT a page layout format people)

    3) NONstandard (ie broswer specific) garbage like conditional comments.

    4) Purposely constructed scripts designed to cause problems.

    In short short of display and some CSS issues you actually have to WORK to
    cause problems with 'valid' HTML. I should note that strickly speaking
    validators will call nonstandard crap like conditional comments 'valid' but
    it you read ALL the specs (appendix as well) you realize that they are
    nonstandard.
    Bruce Grubb, Apr 18, 2004
    #1
    1. Advertising

  2. In article <>,
    Bruce Grubb <> wrote:

    > In article <>,
    > Gregory Weston <> wrote:
    >
    > > In article <>,
    > > Bruce Grubb <> wrote:
    > >
    > > > In article <170420040759517445%>,
    > > > Thomas Reed <> wrote:
    > > >
    > > > > In article <-berlin.de>, Joe Doe
    > > > > <> wrote:
    > > > >
    > > > > > PS: The validator page itself http://validator.w3.org returns 9 errors
    > > > > > about its own html code. "But I read it in the paper!..."
    > > > >
    > > > > A validator really gives you very little useful information except
    > > > > whether your page conforms to standards. Since almost all web browsers
    > > > > render non-standard pages, this is not useful information, IMHO.
    > > >
    > > > Actually it is VERY useful information as by going to the standard rather
    > > > than some browser glitches make is more likely (though not certain) that
    > > > the thing to be read correctly by a 4.01 complient browser you have not
    > > > even heard of.

    > >
    > > A problem arises, however, if an exceedingly popular browser does not
    > > render correct HTML in a worthwhile fashion.

    >
    > Now what 'exceedingly popular browser' does this? Remember HTML is NOT a
    > page layout format (that is what PDF is for) so a specific layout does NOT
    > count.


    a) The comment was provided as an abstract.

    b) Several versions of IE have had problems with various bits of markup.

    > In short short of display and some CSS issues you actually have to WORK to
    > cause problems with 'valid' HTML.


    Um. Wouldn't "display problems" be a shorter way of saying "does not
    render correct HTML in a worthwhile fashion?"

    --
    Standard output is like your butt. Everyone has one. When using a bathroom,
    they all default to going into a toilet. However, a person can redirect his
    "standard output" to somewhere else, if he so chooses. - Jeremy Nixon
    Gregory Weston, Apr 18, 2004
    #2
    1. Advertising

  3. Bruce Grubb

    Bruce Grubb Guest

    In article <>,
    Gregory Weston <> wrote:

    > In article <>,
    > Bruce Grubb <> wrote:
    >
    > > In article <>,
    > > Gregory Weston <> wrote:
    > >
    > > > In article <>,
    > > > Bruce Grubb <> wrote:
    > > >
    > > > > In article <170420040759517445%>,
    > > > > Thomas Reed <> wrote:
    > > > >
    > > > > > In article <-berlin.de>, Joe Doe
    > > > > > <> wrote:
    > > > > >
    > > > > > > PS: The validator page itself http://validator.w3.org returns 9
    > > > > > > errors
    > > > > > > about its own html code. "But I read it in the paper!..."
    > > > > >
    > > > > > A validator really gives you very little useful information except
    > > > > > whether your page conforms to standards. Since almost all web
    > > > > > browsers render non-standard pages, this is not useful
    > > > > > information, IMHO.
    > > > >
    > > > > Actually it is VERY useful information as by going to the standard
    > > > > rather than some browser glitches make is more likely (though
    > > > > not certain) that the thing to be read correctly by a 4.01 complient
    > > > > browser you have not even heard of.
    > > >
    > > > A problem arises, however, if an exceedingly popular browser does not
    > > > render correct HTML in a worthwhile fashion.

    > >
    > > Now what 'exceedingly popular browser' does this? Remember HTML is NOT a
    > > page layout format (that is what PDF is for) so a specific layout does NOT
    > > count.

    >
    > a) The comment was provided as an abstract.
    >
    > b) Several versions of IE have had problems with various bits of markup.


    All of which are what by reading the specs were NONstandard. One of the
    counter arguments has been standard code has far less chance of doing
    something stupid than nonstandard crap.

    > > In short short of display and some CSS issues you actually have to WORK to
    > > cause problems with 'valid' HTML.

    >
    > Um. Wouldn't "display problems" be a shorter way of saying "does not
    > render correct HTML in a worthwhile fashion?"


    No because 'valid' HMTL is more than just shoving the page at a Validator
    and getting no errors. Each HTML spec (2.0, 3.2, 4.01) have certain
    guidelines that need to followed in addition to simple Validation.

    Take the example page given "who uses virtual pc for testing webpages"

    <http://www.goddamn.co.uk/tobyink/scratch/example>

    <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
    <title>Example</title>
    <body><!--[if gte IE 5]><input type example><![endif]--></body>

    Which validates BUT if you read the HTML 4.01 spec you find out it is NOT
    valid HTML. Why? Because HMTL 4.01 supersedes HTML 3.2 and HTML 2.0 and
    defines script as follows:

    It was that as well lamblasting the use of conditional comments. By the
    defintions HTML 4.01 lays out conditional comments ARE scripts:

    "a program that may accompany an HTML document or be embedded directly in
    it. The program executes on the client's machine when the document loads,
    or at some other time such as when a link is activated. HTML's support for
    scripts is independent of the scripting language."
    <http://www.w3.org/TR/REC-html40/interact/scripts.html#h-18.1>

    Conditional Comments are simply a MS was of doing scripts by NOT following
    the guidelines and here syllogism (Aristotlian logic) that proves it:

    1) a program that 'executes' on the client's machine when the
    document loads is a script (defined in HTML4.01 section 18)

    2) conditional comments are commented out commands that
    execute on the client's machine when the document
    is load by a IE browser.

    3) ergo conditional comments ARE scripts reguardless of
    how those who write IE browser supported pages want to handwave it.

    Conditional comments validate simply because under HTML 3.2 there are not
    supposed to be scripts in ANY form and treats the thing as a comment per
    the spec. Its when IE tries to treat the thing as a script that the
    problems occur.

    Also if that was not bad enough there is this warning in the HTML4.01 spec:

    "Some browsers close comments on the first ">" character, so to hide script
    content from such browsers, you can transpose operands for relational and
    shift operators (e.g., use "y < x" rather than "x > y") or use scripting
    language-dependent escapes for ">"."
    < http://www.w3.org/TR/REC-html40/interact/scripts.html>

    Which explains why the above causes some versions of IE to crash.

    Here is what I said in that thread aobut doing this type of stuff:

    If it is a commented out element or the like remove the comments and see if
    the thing still validates; if not then rewrite the thing. Improper
    displaying of elements is not a problem as HTML is NOT nor NEVER will be a
    page markup format.

    Sure you can write infinate loops in any thing HTML calls to and it will
    validate (but will likly cause problems) but that is not the validator's
    fault.

    Validators like HTML authoring program are tools and like any tool they do
    require a little use of common sence. Would you use a wrench to hammer a
    nail; sure you could use it that way but why unless you absolutely had to?

    The same question goes in regards to commented out invalid tags, infinate
    loops, and browser specific crap. Is there a reason it is there other than
    'I want to show that I can write something that validates but will cause
    some browsers to crash' or 'I don't give a flying flip what the WWW uses
    they *must* use browser 'x' version 1.0.1'?

    If there is there is no reason that such junk is there other than an ego
    trip jolly fest than there is NO valid reason for it to be there.

    Also not all validators are created equal. Some like the one at
    webpagesthatsuck <http://www.webpagesthatsuck.com/> are designed to not
    only validate tot he specs but to look for browser specific crap.
    Bruce Grubb, Apr 20, 2004
    #3
  4. Bruce Grubb

    Spartanicus Guest

    Bruce Grubb <> wrote:

    >'valid' HMTL is more than just shoving the page at a Validator
    >and getting no errors.


    Incorrect, valid means just that, nothing more, nothing less. It's just
    that validity doesn't mean that much.

    >Also not all validators are created equal.


    They are, or they are not validators.

    --
    Spartanicus
    Spartanicus, Apr 20, 2004
    #4
  5. Bruce Grubb wrote:

    > All of which are what by reading the specs were NONstandard. One of the
    > counter arguments has been standard code has far less chance of doing
    > something stupid than nonstandard crap.


    That I have certainly never disputed.

    > No because 'valid' HMTL is more than just shoving the page at a Validator
    > and getting no errors.


    No -- you're confusing "valid HTML" with "good HTML". "Good HTML" is more
    than just shoving the page at a Validator and getting no errors. Shoving
    the page at a Validator and getting no errors is perfectly sufficient for
    valid HTML.

    > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
    > <title>Example</title>
    > <body><!--[if gte IE 5]><input type example><![endif]--></body>
    >
    > Which validates BUT if you read the HTML 4.01 spec you find out it is NOT
    > valid HTML. Why? Because HMTL 4.01 supersedes HTML 3.2 and HTML 2.0 and


    HTML 4.01 is a later standard, though HTML 3.2 and 2.0 are not obsoleted
    by it.

    Arguably HTML 2.0 is more of a standard than HTML 4.01 is: HTML 2.0 is an
    Internet Engineering Task Force (IETF) Standard whereas HTML 4.01 is only
    a W3C Recommendation.

    > defines script as follows:


    Who cares? For the last time, THE EXAMPLE ABOVE DOES NOT CONTAIN ANY
    SCRIPT DATA!

    > Conditional Comments are simply a MS was of doing scripts by NOT
    > following the guidelines and here syllogism (Aristotlian logic) that
    > proves it:
    >
    > 1) a program that 'executes' on the client's machine when the
    > document loads is a script (defined in HTML4.01 section 18)


    Agreed.

    > 2) conditional comments are commented out commands that
    > execute on the client's machine when the document is load by a IE
    > browser.


    Refuted. Conditional comments do not "execute".

    > 3) ergo conditional comments ARE scripts reguardless of
    > how those who write IE browser supported pages want to handwave it.


    Based on a false assumption in #2.

    > Conditional comments validate simply because under HTML 3.2 there are
    > not supposed to be scripts


    They also validate under HTML 2.0, HTML 4.01, XHTML 1.0, XHTML 1.1 and
    provisional drafts of XHTML 2.0, four of which are allowed to contain
    scripts (not that scripting has any bearing on this case).

    > "Some browsers close comments on the first ">" character, so to hide
    > script content from such browsers, you can transpose operands for
    > relational and shift operators (e.g., use "y < x" rather than "x > y")
    > or use scripting language-dependent escapes for ">"." <
    > http://www.w3.org/TR/REC-html40/interact/scripts.html>
    >
    > Which explains why the above causes some versions of IE to crash.


    No it doesn't. The crash is cause by the following few bytes alone:

    <input type example>

    The conditional comment is just used to force it to validate.

    > 'I want to show that I can write something that
    > validates but will cause some browsers to crash'


    My example was made entirely to prove that exact point.

    However, some people refuse to accept it in the face of irrefutable facts.

    > Also not all validators are created equal. Some like the one at
    > webpagesthatsuck <http://www.webpagesthatsuck.com/> are designed to not
    > only validate tot he specs but to look for browser specific crap.


    Then they're not validators -- they're linters or what have you.

    *plonk*

    --
    Toby A Inkster BSc (Hons) ARCS
    Contact Me - http://www.goddamn.co.uk/tobyink/?page=132
    Toby A Inkster, Apr 20, 2004
    #5
  6. Bruce Grubb

    C A Upsdell Guest

    "Spartanicus" <> wrote in message
    news:...
    > Bruce Grubb <> wrote:
    >
    > >'valid' HMTL is more than just shoving the page at a Validator
    > >and getting no errors.

    >
    > Incorrect, valid means just that, nothing more, nothing less. It's just
    > that validity doesn't mean that much.


    Validity means three things: fewer surprises; less time debugging; and one
    less place to look if pages don't work.

    > >Also not all validators are created equal.

    >
    > They are, or they are not validators.


    Some have clearer error messages.
    C A Upsdell, Apr 20, 2004
    #6
  7. In article <>,
    Toby A Inkster <> wrote:

    > > No because 'valid' HMTL is more than just shoving the page at a Validator
    > > and getting no errors.

    >
    > No -- you're confusing "valid HTML" with "good HTML". "Good HTML" is more
    > than just shoving the page at a Validator and getting no errors. Shoving
    > the page at a Validator and getting no errors is perfectly sufficient for
    > valid HTML.


    Most languages, and I doubt HTML is an exception (especially when
    scripting is involved), have both static and dynamic validity
    requirements. Validators can generally only check the static features
    of the page, and perhaps use some heuristics to try to check for some
    dynamic errors.

    Also, the validator programmer may not have thought to check for all the
    possible errors that can appear.

    Assuming that a page that successfully passes a validator is truly valid
    would be like assuming that a C program that compiles with no errors is
    bug-free.

    --
    Barry Margolin,
    Arlington, MA
    Barry Margolin, Apr 20, 2004
    #7
  8. On Tue, 20 Apr 2004 17:24:59 -0400, Barry Margolin wrote:

    > Most languages, and I doubt HTML is an exception (especially when
    > scripting is involved), have both static and dynamic validity
    > requirements. Validators can generally only check the static features
    > of the page, and perhaps use some heuristics to try to check for some
    > dynamic errors.


    I don't know what version of HTML you're using, but mine doesn't have any
    dynamic elements.
    Brendan Taylor, Apr 20, 2004
    #8
  9. In article <ozhhc.179115$Pk3.163855@pd7tw1no>,
    Brendan Taylor <> wrote:

    > On Tue, 20 Apr 2004 17:24:59 -0400, Barry Margolin wrote:
    >
    > > Most languages, and I doubt HTML is an exception (especially when
    > > scripting is involved), have both static and dynamic validity
    > > requirements. Validators can generally only check the static features
    > > of the page, and perhaps use some heuristics to try to check for some
    > > dynamic errors.

    >
    > I don't know what version of HTML you're using, but mine doesn't have any
    > dynamic elements.


    Even when there are scripts in the HTML?

    --
    Barry Margolin,
    Arlington, MA
    Barry Margolin, Apr 21, 2004
    #9
  10. Bruce Grubb

    Mark Parnell Guest

    On Wed, 21 Apr 2004 00:07:57 -0400, Barry Margolin <>
    declared in alt.html,comp.sys.mac.apps:

    > Even when there are scripts in the HTML?


    There aren't scripts in any HTML. There is a script element that allows
    you to embed scripts, which are written in languages other than HTML.
    HTML is a markup language - it cannot _do_ anything. It just describes
    the structure of the document.

    --
    Mark Parnell
    http://www.clarkecomputers.com.au
    Mark Parnell, Apr 21, 2004
    #10
  11. In article <qbngzymyn70a$.1rtz60hzjpjk1$>,
    Mark Parnell <> wrote:

    > On Wed, 21 Apr 2004 00:07:57 -0400, Barry Margolin <>
    > declared in alt.html,comp.sys.mac.apps:
    >
    > > Even when there are scripts in the HTML?

    >
    > There aren't scripts in any HTML. There is a script element that allows
    > you to embed scripts, which are written in languages other than HTML.
    > HTML is a markup language - it cannot _do_ anything. It just describes
    > the structure of the document.


    IIRC, the problems the OP was having were with Javascript scripts
    embedded on the web page. So if you don't include the validity of the
    scripts as part of the validity of the document itself, what was the
    point of bringing up the validitation issue in the first place?

    --
    Barry Margolin,
    Arlington, MA
    Barry Margolin, Apr 21, 2004
    #11
  12. Bruce Grubb

    Mark Parnell Guest

    On Wed, 21 Apr 2004 01:21:02 -0400, Barry Margolin <>
    declared in alt.html,comp.sys.mac.apps:

    > IIRC, the problems the OP was having were with Javascript scripts
    > embedded on the web page. So if you don't include the validity of the
    > scripts as part of the validity of the document itself, what was the
    > point of bringing up the validitation issue in the first place?


    I don't subscribe to comp.sys.mac.apps, so I came in halfway through
    this thread (when it was crossposted to alt.html), and validation was
    already part of the discussion.

    Looking at the beginning of the thread, it seems that validation was
    brought up as a separate example of clueless programmers. While I agree
    that to some extent not coding HTML to standards is a different issue to
    that of not coding Javascript to standards (which was the OP's problem),
    the two are obviously related (they're both about coding to standards),
    and as such I think it was relevant to the discussion.

    Besides, this is Usenet. Since when does there have to be a point? :-D

    --
    Mark Parnell
    http://www.clarkecomputers.com.au
    Mark Parnell, Apr 21, 2004
    #12
  13. Barry Margolin wrote:

    > Validators can generally only check the static features
    > of the page, and perhaps use some heuristics to try to check for some
    > dynamic errors.


    Validators can and should only check that a page conforms to the DTD.

    While a page may contain other errors (links to non-existing URLs,
    speling mistakes, scripting problems, etc), as long as it conforms to the
    DTD it is *valid*.

    Hence my point about the difference between "valid HTML" and "good HTML".
    (Usually the former is a subset of the latter.)

    --
    Toby A Inkster BSc (Hons) ARCS
    Contact Me - http://www.goddamn.co.uk/tobyink/?page=132
    Toby A Inkster, Apr 21, 2004
    #13
  14. Bruce Grubb

    Philo D Guest

    In article <>, Toby A
    Inkster <> wrote:

    >
    > [...] speling mistakes [...]


    gud wun
    Philo D, Apr 21, 2004
    #14
  15. In article <>,
    Toby A Inkster <> wrote:

    > Barry Margolin wrote:
    >
    > > Validators can generally only check the static features
    > > of the page, and perhaps use some heuristics to try to check for some
    > > dynamic errors.

    >
    > Validators can and should only check that a page conforms to the DTD.
    >
    > While a page may contain other errors (links to non-existing URLs,
    > speling mistakes, scripting problems, etc), as long as it conforms to the
    > DTD it is *valid*.
    >
    > Hence my point about the difference between "valid HTML" and "good HTML".
    > (Usually the former is a subset of the latter.)


    Right. So being valid may be a necessary qualification (I say "may"
    because most browsers are forgiving of some errors), but not a
    sufficient qualification. Passing a validator is just a first step in
    deciding whether a page is any good. And failing a validation check is
    a good indicator that the designer doesn't know what he's doing.
    Although these days I suspect the blame more likely falls on the
    designers of the automated tools for generating web pages -- how many
    people still hand-code their HTML?

    --
    Barry Margolin,
    Arlington, MA
    Barry Margolin, Apr 21, 2004
    #15
  16. Bruce Grubb

    Dave Hinz Guest

    On Wed, 21 Apr 2004 10:46:34 -0400, Barry Margolin <> wrote:

    > Although these days I suspect the blame more likely falls on the
    > designers of the automated tools for generating web pages -- how many
    > people still hand-code their HTML?


    /me raises hand

    Of course, my HTML looks, well, hand-coded. Content over flash and all
    that.

    I think the point that has got lost in the weeds of the "valid vs. good"
    argument is that the website this guy was complaining about, is
    written poorly and insufficiently validated. I'll bet 20 bucks that
    it was only tested on IE on Windows, with no thought (or maybe even,
    no awareness) of other browsers / platforms. To me, that's a message
    that they don't understand their consumer base and an invitation to
    go spend my money elsewhere. I've actually switched banks over crappy
    website issues... not only did they want me to be running IE, but they
    checked for IE 5.01 as an exact match, not "or newer". Sure, I can have
    Opera or whatever identify itself as that, but, c'mon guys...

    Bad HTML is out there. Validators will find some of it. If your
    site doesn't work, _and_ it doesn't validate, then you missed two chances
    to do it right.

    Dave Hinz
    Dave Hinz, Apr 21, 2004
    #16
  17. Barry Margolin <> wrote in message news:<>...
    > Although these days I suspect the blame more likely falls on the
    > designers of the automated tools for generating web pages -- how many
    > people still hand-code their HTML?


    I do...

    --
    Dan
    Daniel R. Tobias, Apr 21, 2004
    #17
  18. Bruce Grubb

    Peter KERR Guest

    > > > Even when there are scripts in the HTML?
    > >
    > > There aren't scripts in any HTML. There is a script element that allows
    > > you to embed scripts, which are written in languages other than HTML.
    > > HTML is a markup language - it cannot _do_ anything. It just describes
    > > the structure of the document.

    >
    > IIRC, the problems the OP was having were with Javascript scripts
    > embedded on the web page. So if you don't include the validity of the
    > scripts as part of the validity of the document itself, what was the
    > point of bringing up the validitation issue in the first place?
    >


    Validators (mostly) check the HTML, a straightforward structural, syntax
    and spelling check.

    Whose gonna validate the Tower of Babel of scripting languages?
    Against what shifting sands of provisional specs?
    Heck, the OP had an inkling, because some browsers don't support
    properly some scripting languages. They shouldn't have to if the HTML is
    clean, and the scripting is all serverside, but this is the real world
    :-(
    Peter KERR, Apr 21, 2004
    #18
  19. Bruce Grubb

    Mark Parnell Guest

    On Wed, 21 Apr 2004 10:46:34 -0400, Barry Margolin <>
    declared in alt.html,comp.sys.mac.apps:

    > how many people still hand-code their HTML?


    You'll find that most of the regulars here do.

    --
    Mark Parnell
    http://www.clarkecomputers.com.au
    Mark Parnell, Apr 21, 2004
    #19
  20. Bruce Grubb

    Mark Parnell Guest

    On Thu, 22 Apr 2004 08:59:15 +1000, Mark Parnell
    <> declared in
    alt.html,comp.sys.mac.apps:

    > You'll find that most of the regulars here do.


    "Here" being alt.html. Forgot about the crosspost. :)

    --
    Mark Parnell
    http://www.clarkecomputers.com.au
    Mark Parnell, Apr 22, 2004
    #20
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Wipkip

    A very sad ? about frames

    Wipkip, Oct 15, 2003, in forum: HTML
    Replies:
    8
    Views:
    355
  2. Bruce Grubb
    Replies:
    2
    Views:
    445
    Gregory Weston
    Apr 22, 2004
  3. JustBoo

    Wow, This Is So Sad...

    JustBoo, Jul 22, 2009, in forum: C++
    Replies:
    17
    Views:
    548
    JustBoo
    Jul 27, 2009
  4. Zac Burns

    Why is python so sad?

    Zac Burns, Oct 14, 2009, in forum: Python
    Replies:
    0
    Views:
    254
    Zac Burns
    Oct 14, 2009
  5. Mike Hunter

    Sad experience with redirecting to /dev/null

    Mike Hunter, Apr 2, 2004, in forum: Perl Misc
    Replies:
    9
    Views:
    157
    Joe Smith
    Apr 4, 2004
Loading...

Share This Page