Inconsistent results across browsers

[Dudely]

My web page displays just fine under IE7, but under IE6 about 90% of
it is just plain missing. I get the top of the page, and the bottom
of the page... but not the middle. I have not tested with any other
browsers.

The page is at www.greengoldcapital.com

I use considerable amounts of CSS, and a lot of other things including
javascript, SSI, PHP, mySQL, and a partridge in a pear tree. I have a
third party application integrated in, as well as "tips & tricks" I'm
using from others, some of which are way over my head; especially the
menu.

I'm at a loss as to where to even start looking for the problem. I
was thinking maybe it was a css issue, yet I use css on the part that
shows up too. As far as I can tell, there's nothing especially
different about what shows up versus what doesn't show up... so I'm
stumped at this point.

Thank you for any good ideas.

[Jim Moe]

Dudely wrote:

My web page displays just fine under IE7, but under IE6 about 90% of
it is just plain missing. I get the top of the page, and the bottom
of the page... but not the middle. I have not tested with any other
browsers.

The page is at www.greengoldcapital.com

It is mostly blank in Firefox v2, too.
Correct all of those validation errors first.
<http://validator.w3.org/check?verbose=1&uri=http%3A%2F%2Fwww.greengoldcapi tal.com%2FOR%2Findex.php>

--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)

[Bergamot]

Dudely wrote:

My web page displays just fine under IE7, but under IE6 about 90% of
it is just plain missing.

www.greengoldcapital.com

In Seamonkey/Firefox, all the content is off the screen on the right.
Scroll over to see it. That should give some indication of one place to
look. As for IE6, the content shows up when JS is disabled. That should
give you another pointer.

I use considerable amounts of CSS, and a lot of other things including
javascript, SSI, PHP, mySQL, and a partridge in a pear tree.

I'm at a loss as to where to even start looking for the problem.

So am I. You need to simplify it if you expect to debug it, but the
mozilla DOM inspector may be helpful.

BTW, IE6 has lots of bugs with floats. See
http://www.positioniseverything.net/explorer.html

--
Berg

[Dudely]

On Jun 1, 12:47 pm, Jim Moe <[e-mail]:

Dudelywrote:

My web page displays just fine under IE7, but under IE6 about 90% of
it is just plain missing. I get the top of the page, and the bottom
of the page... but not the middle. I have not tested with any other
browsers.

The page is atwww.greengoldcapital.com

It is mostly blank in Firefox v2, too.
Correct all of those validation errors first.
<http://validator.w3.org/check?verbose=1&uri=http%3A%2F%2Fwww.greengol...>

--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)

Thanks for the pointer. I had been using my browser's built-in
validator, but it missed a bunch of things that w3 picked up. On the
other hand, while I was editing my files, I accidentely deleted the
<htmltag, but w3 validator failed to notice, whereas my editor's
validator did notice. In any event, it now validates.

Once I got it all validated, I then went to test it under IE6.

Amazingly enough, it also displayed under IE6. Then I made a very
small change, that broke IE6 again.

The change I made, had to do with the following:

<!--[if gt IE 6]>

I was forced to put in the above statement, in order to make the page
validate.
The above statement commented out the menu, which uses a .htc file.

Because everything displays except the menu, under IE6 I decided to
change the 6 to a 5, to confirm it worked in IE6, which is when it
broke. So I put it back to 6 and of course it works again.

So, we seem to have narrowed the problem down.

The file wouldn't validate originally because it didn't like how the
<ULtags were nested, which is pretty much essential to how the menu
works. The menu, was written by someone else, and it's pretty far
over my head. Quite frankly, I don't understand how most of it works,
but it does under IE7, as does the rest of the website.

So, unless I re-do the menu in some totally different way, I'm at a
loss as to how to fix it. The menu took me hours upon hours upon
hours to get working under IE7, so I'm really reluctant to just throw
it away, unless there's something that looks & functions essentially
the same.

Good ideas?

Thank you

[Jim Moe]

Dudely wrote:

>
The file wouldn't validate originally because it didn't like how the
<ULtags were nested, which is pretty much essential to how the menu
works. The menu, was written by someone else, and it's pretty far
over my head. Quite frankly, I don't understand how most of it works,
but it does under IE7, as does the rest of the website.

Getting a page to work in one browser, especially one that is not widely
used yet, is not much use.
The error message is very clear about what is wrong. But just to be nice...
Change this:
<ul>
<li>Bleh</li>
<ul>
<li>yadda</li>
</ul>
</ul>

to:
<ul>
<li>Bleh
<ul>
<li>yadda</li>
</ul>
</li>
</ul>
--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)

[JD]

Dudely wrote:

My web page displays just fine under IE7, but under IE6 about 90% of
it is just plain missing. I get the top of the page, and the bottom
of the page... but not the middle. I have not tested with any other
browsers.

The page is at www.greengoldcapital.com

I use considerable amounts of CSS, and a lot of other things including
javascript, SSI, PHP, mySQL, and a partridge in a pear tree. I have a
third party application integrated in, as well as "tips & tricks" I'm
using from others, some of which are way over my head; especially the
menu.

I'm at a loss as to where to even start looking for the problem. I
was thinking maybe it was a css issue, yet I use css on the part that
shows up too. As far as I can tell, there's nothing especially
different about what shows up versus what doesn't show up... so I'm
stumped at this point.

Thank you for any good ideas.

Hello

For such a simple design, your code is mind-bogglingly complicated. The
best suggestion I can offer is that you start from scratch, only this
time you:

1. Keep things simple
2. Make sure you understand what you're putting in
3. Test your page in a variety of browsers /as you build it/
4. Keep things simple

And did I mention that you should keep things simple?

If starting from scratch is not possible for some reason, you could try
scrapping the menu, which will probably go a long way towards
simplifying the page and solving some problems. You mentioned that
you're unwilling to scrap the menu because it was hard to get working
but that's not a good reason to keep it. Your only consideration should
be: Do you really need a drop down menu?

JD

[Dudely]

On Jun 1, 10:50 pm, Jim Moe <[e-mail]:

Dudelywrote:

The file wouldn't validate originally because it didn't like how the
<ULtags were nested, which is pretty much essential to how the menu
works. The menu, was written by someone else, and it's pretty far
over my head. Quite frankly, I don't understand how most of it works,
but it does under IE7, as does the rest of the website.

Getting a page to work in one browser, especially one that is not widely
used yet, is not much use.
The error message is very clear about what is wrong. But just to be nice...
Change this:
<ul>
<li>Bleh</li>
<ul>
<li>yadda</li>
</ul>
</ul>

to:
<ul>
<li>Bleh
<ul>
<li>yadda</li>
</ul>
</li>
</ul>

Now that you've said it, it seems obvious. But until you said it, I
didn't see it. Thank you. I've made those changes and it still works
in IE7, plus now it validates.

That led me to some more experimentation, and I think I have now
isolated it to the css declarations for the menu.

However, in the process of making the above mentioned changes, I
partially broke the background coloring for some of the top level menu
items. At this point I'm too tired to think straight, I'll come back
to it tommorrow.

In addition, prior to make those changes, IE6 was displaying the menu,
though none of the dropdown functionality worked. Now, the menu is
not displaying again on IE6, though the rest of the page is still
there, so I guess progress has been made.

Thanks again

[Dudely]

On Jun 1, 1:54 pm, Bergamot <[e-mail]:

Dudelywrote:

My web page displays just fine under IE7, but under IE6 about 90% of
it is just plain missing.

www.greengoldcapital.com

In Seamonkey/Firefox, all the content is off the screen on the right.
Scroll over to see it. That should give some indication of one place to
look. As for IE6, the content shows up when JS is disabled. That should
give you another pointer.

I use considerable amounts of CSS, and a lot of other things including
javascript, SSI, PHP, mySQL, and a partridge in a pear tree.

I'm at a loss as to where to even start looking for the problem.

So am I. You need to simplify it if you expect to debug it, but the
mozilla DOM inspector may be helpful.

BTW, IE6 has lots of bugs with floats. Seehttp://www.positioniseverything.net/explorer.html

I turned off one of my floats, and the menu showed up on IE6, though
it still doesn't drop down properly. It doesn't seem to have had a
bad effect in IE7 as far as I can tell, but I need sleep.

I'll play some more with this tommorrow.

Thank you for the pointer.

[rf]

"JD" <us**@example.netwrote in message
news:5c*************@mid.individual.net...

Dudely wrote:

>The page is at www.greengoldcapital.com

For such a simple design, your code is mind-bogglingly complicated. The
best suggestion I can offer is that you start from scratch, only this time
you:

Agreed. Could be done with a tenth of the code.

>
1. Keep things simple
2. Make sure you understand what you're putting in

2.5 and only put it in if you are *very very sure* you need it.

3. Test your page in a variety of browsers /as you build it/

3.5 but keep the code to the specifications. Do not use browser specific
code.

4. Keep things simple

Dudely, you are approaching this from the wrong angle. You appear to be
jumping through all sorts of hoops to make sure it works in various flavours
of IE, even to the point of sniffing browsers and sniffing different IE
versions.

Don't.

Code it to the specifications with the absolute minimum of code. No
javascript at all and an absolute minimum of CSS, just to get the content
and the layout correct. Then gradually add the eye candy stuff (still not
using javascript), testing in all the browsers you have to hand after each
step.

Only then may you put in some javascript to *add*[1] to the page if you
wish, keeping in mind that a significant minority do not have javascript.

[1] FWIW your page looks *MUCH* better in FF with javascript turned off :-)

--
Richard.

[Chaddy2222]

rf wrote:

"JD" <us**@example.netwrote in message
news:5c*************@mid.individual.net...

Dudely wrote:

The page is at www.greengoldcapital.com

For such a simple design, your code is mind-bogglingly complicated. The
best suggestion I can offer is that you start from scratch, only this time
you:

Agreed. Could be done with a tenth of the code.

1. Keep things simple
2. Make sure you understand what you're putting in

2.5 and only put it in if you are *very very sure* you need it.

Yes, I could not agree more, the amount of large sites I have noticed
with things that arn't needed is amaizing.

3. Test your page in a variety of browsers /as you build it/

3.5 but keep the code to the specifications. Do not use browser specific
code.

Yeah and that goes for you people developing online applications as
well.

4. Keep things simple

Dudely, you are approaching this from the wrong angle. You appear to be
jumping through all sorts of hoops to make sure it works in various flavours
of IE, even to the point of sniffing browsers and sniffing different IE
versions.

Don't.

Code it to the specifications with the absolute minimum of code. No
javascript at all and an absolute minimum of CSS, just to get the content
and the layout correct. Then gradually add the eye candy stuff (still not
using javascript), testing in all the browsers you have to hand after each
step.

Only then may you put in some javascript to *add*[1] to the page if you
wish, keeping in mind that a significant minority do not have javascript.

[1] FWIW your page looks *MUCH* better in FF with javascript turned off :-)

I agree with all this and also remember to use a container div to make
sure that things line up correctly in a larger number of browsers, I
have ran into difficulties before when not useing some kind of
wrapper around the entire page.
--
Regards Chad. http://freewebdesign.awardspace.biz

[Andy Dingley]

On 1 Jun, 20:29, Dudely <[e-mail]:

My web page displays just fine under IE7, but under IE6 about 90% of
it is just plain missing.

Don't EVER use this:
if ((navigator.appVersion.indexOf("MSIE") 0)
&& (parseInt(navigator.appVersion) <= 6)) {

Don't EVER use this:
<!--[if lt IE 7]>

You might use this, but I strongly advise against it.
<!--[if gt IE 7]>

If you must use IE conditionals, then the only safe form to use is
this one:
<!--[if IE]>

The reasoning is to do with the "fall back" behaviour in browsers that
correctly ignore this M$oft-specific extension.
Some general rules on web design (in order)

* Write standards-based code that passes HTML and CSS validation.

* Adjust the design until you're happy that it works in a standards-
compliant browser. I suggest Firefox and the Marc Geuery HTML
validator extension.

* Keep checking that it's _still_ valid!

* See how badly the broken IE mangles it. Fix, work-around or avoid
particular pieces of the standard that IE can't cope with.

* Use the <!--[if IE]form to fix anything you can't possible fix by
other means.

* Don't use an IE-only feature unless you absolutely _must_, and
you're already experienced enough to do valid standards-based design
and coding for other standards-based browsers.

* Never use an "If not IE" test. They fail outside IE, so always use
the "If IE" form instead.

* Never test for specific IE versions. This gives non-robust code that
costs a lot in maintenance when IE changes.

* NEVER break the validation, the standards-based code, or the
behaviour in other browsers just to try and keep IE happy.

[Jukka K. Korpela]

Scripsit Andy Dingley:

Don't EVER use this:
if ((navigator.appVersion.indexOf("MSIE") 0)
&& (parseInt(navigator.appVersion) <= 6)) {

If you do browser sniffing regarding IE, that trick works roughly as well as
the alternatives. I don't think it's particular useful to nitpick on it.

Don't EVER use this:
<!--[if lt IE 7]>

You sound very dogmatic, but in a pointless way. You can declare all browser
sniffing as wrong, and that would be a defendable though a bit too
idealistic viewpoint. But if you approve browser sniffing at all, surely you
need to recognize the fundamental difference between IE 7 and previous
versions of IE.

If you must use IE conditionals, then the only safe form to use is
this one:
<!--[if IE]>

Sorry, but that's not only ridiculous - it will also confuse newbies if you
take you seriously.

The reasoning is to do with the "fall back" behaviour in browsers that
correctly ignore this M$oft-specific extension.

There's no "fall back". They simply just read the style sheet by the book
and ignore the part that they are expected to ignore, no matter whether you
have [if IE] or [if lt IE 7].

* Never use an "If not IE" test.

What strawman is that?

* Never test for specific IE versions. This gives non-robust code that
costs a lot in maintenance when IE changes.

You're completely wrong. If your page works fine on browsers other than IE
_and_ on IE 7, it would be foolish to tell IE 7 to use the tricks that you
have written to make IE 6 behave.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

[rf]

"Andy Dingley" <di*****@codesmiths.comwrote in message
news:11*********************@q69g2000hsb.googlegro ups.com...

* Never test for specific IE versions. This gives non-robust code that
costs a lot in maintenance when IE changes.

Disagree.

<!--[if IE 6]allows me to inject a bit of javascript skulduggery that
makes {position: fixed} sort of work for those still using this specific
browser. (javascript *must* be enabled to use the page anyway). Same with
the javascript solution to CSS dropdown menus in IE6 although I also employ
a non-javascript degredation.

No other browser will see this javascript. I have nothing to 'maintain' in
the future, other than to someday remove the code when IE6 dies.

--
Richard.

[Andy Dingley]

On 4 Jun, 12:14, "Jukka K. Korpela" <[e-mail]:

You sound very dogmatic, but in a pointless way.

Do _you_ really expect a reply to that?

[Andy Dingley]

On 4 Jun, 12:35, "rf" <[e-mail]:

<!--[if IE 6]allows me to inject a bit of javascript skulduggery that
makes {position: fixed} sort of work for those still using this specific
browser.

This sort of "If old version" test is certainly more stable than
something using "If new version". That's the real problem, and the
one that needs to be guarded against carefully.

If there _is_ an old version though, why go to this extra trouble to
support it? It's already an "old" version that's withering away, or
else it's not yet safe enough to treat it this way. If so, why embed
what's becoming obsolete cruft into your shiny new pages?

It's not 1999 any more, browser compatibility isn't still so bad that
you're forced into supporting browser vagaries because it's the only
way to achieve a useful page.

[Jukka K. Korpela]

Scripsit Andy Dingley:

On 4 Jun, 12:14, "Jukka K. Korpela" <[e-mail]:

>You sound very dogmatic, but in a pointless way.

Do _you_ really expect a reply to that?

Of course not. It was not a question.

Thank you for confirming my observations by this posting of yours, where you
did not make any comment on the substance and just wasted bandwidth and
everyone's time with a foolish question. You confirmed my observations,
since you are known to be an intelligent person, who would surely have
presented rational counterarguments if he had one.

We conclude that <!--[if lt IE 7]... <![endif]and similar pseudocomments
are a good way of dealing with problems in IE, _if_ an authors decides to
use browser sniffing.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

[Sherm Pendley]

Andy Dingley <di*****@codesmiths.comwrites:

* Never use an "If not IE" test. They fail outside IE

I never use IE conditionals myself, but I don't understand this advice.
Isn't the whole point of an "if not IE" test that it will fail outside IE?
IE will parse the conditional and find it negative. Other browsers will
ignore the conditional, so they will render its contents, same as if
they'd parsed it as positive.

sherm--

--
Web Hosting by West Virginians, for West Virginians: http://wv-www.net
Cocoa programming in Perl: http://camelbones.sourceforge.net

[Andy Dingley]

On Mon, 04 Jun 2007 15:36:36 -0400, Sherm Pendley <sp******@dot-app.org>
wrote:

>Andy Dingley <di*****@codesmiths.comwrites:

>* Never use an "If not IE" test. They fail outside IE

I never use IE conditionals myself, but I don't understand this advice.
Isn't the whole point of an "if not IE" test that it will fail outside IE?

No, that's the "If IE" form. Non-IE browsers then see them as a coment.
Your valid argument is the justification for my practice.