Multiple Hacks and 3 pixel jog

[R0bert Neville]

I am having major issues in IE. This web development exercise involves
learning why certain hacks work and others don't. My web page renders
ok in Firefox.

The main problem in my web page relates to complexity; emulating
position fixed in IE. If I simplify the page, the issues subside,
which does not help me learn about web development idiosyncrasies.

The height %1 hack did not alleviate the situation either. Look at the
CSS, simple.css, for details.

The three pixel jog presents a major problem on a long page. The
information at PIE was helpful, yet did not help me resolve the
situation. Again, the main inquiry refers to managing multiple hacks
and pinpointing issues within IE. The DOM inspector in Firefox helps,
yet I have not found its equivalent in IE.
http://www.positioniseverything.net/...reepxtest.html

Do I have to apply this height hack or dimension hack (hasLayout) to
every element on the web page including <p> elements?

The following page is pared down examples for those purists, who
prefer a simplified page.

http://neville.f2o.org/Template_TEMP1.html

Full page
http://neville.f2o.org/Template_noXML.html

[Els]

R0bert Neville wrote:

I am having major issues in IE. This web development exercise involves
learning why certain hacks work and others don't. My web page renders
ok in Firefox.

The main problem in my web page relates to complexity; emulating
position fixed in IE. If I simplify the page, the issues subside,
which does not help me learn about web development idiosyncrasies.
Of course it does. Troubleshooting is exactly that - take styles
and/or elements off until the problem disappears. Then deduct which
element and/or style caused the problem.

[3px jog] Do I have to apply this height hack or dimension hack (hasLayout) to
every element on the web page including <p> elements?
The 1%height hack only needs to be applied to the containing element,
which in most cases is the content div.
The following page is pared down examples for those purists, who
prefer a simplified page.

http://neville.f2o.org/Template_TEMP1.html
If that is simplified...
Full page
http://neville.f2o.org/Template_noXML.html

Okay, I can see how this page is even fuller than the other one, but
the simple one is by no means simplified. (imo of course :-) )

--
Els http://locusmeus.com/
Sonhos vem. Sonhos vão. O resto é imperfeito.
- Renato Russo -

[Spartanicus]

R0bert Neville <ro***************@yahoo.com> wrote:

I am having major issues in IE. This web development exercise involves
learning why certain hacks work and others don't. My web page renders
ok in Firefox.

The main problem in my web page relates to complexity; emulating
position fixed in IE. If I simplify the page, the issues subside,
which does not help me learn about web development idiosyncrasies.
The situation should have taught you that it's unwise to over complicate
things in an attempt to get IE to do stuff that is beyond it's natural
ability. Keeping an element in view with position:fixed should never be
a requirement for a page to work, just let it scroll in browsers that
don't support it.
http://neville.f2o.org/Template_noXML.html

Some other tips:

A document should contain only one level 1 header, not 13.

<a name="top" id="top"></a> is old school coding only useful if NS4
compatibility is a must. <h2 id="top">Definition List Selectors
Example</h2> is the modern way.

Don't rely on javascript for anything essential, avoid non standard
Ecma/javascript, use capability detection.

Use a width that scales with the text size for layout containers with
text, otherwise your site will break if the user zooms the text.

Sizing fonts down to 62.5% of my configured preference results in
unreadable text.

There are more browsers than IE and Mozilla, a site published on the www
should be constructed to work on most of them.

Don't fix the width of the layout, allow it to flow in the width of the
available viewport.

--
Spartanicus

[R0bert Neville]

On Mon, 02 May 2005 10:14:34 GMT, Spartanicus
<in*****@invalid.invalid> wrote:

R0bert Neville <ro***************@yahoo.com> wrote: Some other tips:
A document should contain only one level 1 header, not 13.
What are you saying by this statement? Use <H2> instead after the
first level one. Interesting point if that's what you mean. Please
clarify and include backup reference.
<h2 id="top">Definition List Selectors Example</h2> is the modern way.
Cool. I like this approach better.
Don't rely on javascript for anything essential, avoid non standard
Ecma/javascript, use capability detection.
Bah Humbug! Javascript is essential to this document template
solution.
Use a width that scales with the text size for layout containers with
text, otherwise your site will break if the user zooms the text. Sizing fonts down to of my configured preference results in
unreadable text.
This approach assumes the default browser text is about 16px. Thus
62.5% brings it to 10px, and then it scales up to 12px in the main
container. If your resolution is so high, then you probably adjust the
text size on most sites anyway. At least, my text scales.
There are more browsers than IE and Mozilla, a site published on the www
should be constructed to work on most of them.
The solution is client-based. The page is only up on a server for
debugging. IE support is only relevant in debugging terms.
Don't fix the width of the layout, allow it to flow in the width of the
available viewport.

The MainContent container has a max-width and min-width for browsers
that follow CSS 2 specifications. I am currently working on a
max-width solution in IE. In addition, the simplified page took a
fixed width approach for the purist.

[Jan Roland Eriksson]

On Mon, 02 May 2005 14:16:06 -0700, R0bert Neville
<ro***************@yahoo.com> wrote:

On Mon, 02 May 2005 10:14:34 GMT, Spartanicus
<in*****@invalid.invalid> wrote:

R0bert Neville <ro***************@yahoo.com> wrote:Some other tips:

A document should contain only one level 1 header, not 13. What are you saying by this statement? Use <H2> instead after the
first level one. Interesting point if that's what you mean. Please
clarify and include backup reference.

Have a look at the ISO version of HTML to find out why.

<https://www.cs.tcd.ie/15445/15445.html>

[ISO = International Standards Organisation, officially supported by
just about as many countries that you can find as members of the UN]

[...]
Don't rely on javascript for anything essential, avoid non standard
Ecma/javascript, use capability detection.

Bah Humbug! Javascript is essential to this document template
solution.

You are losing me there.

How on earth are you going to create a web site that will convey the
info you want to give me, if you are at the same time saying that in
order to do so you want to execute _your_ program code on _my_ system?

Forget that, I'm not interested, my system is under my control, not
yours.

--
Rex

[Spartanicus]

R0bert Neville <ro***************@yahoo.com> wrote:

A document should contain only one level 1 header, not 13.
What are you saying by this statement? Use <H2> instead after the
first level one. Interesting point if that's what you mean. Please
clarify and include backup reference.

A page needs a brief description of it's entire content, this should be
provided by a level 1 header. From that follows that multiple level 1
headers is incorrect markup. Search the web if you need references.

Don't rely on javascript for anything essential, avoid non standard
Ecma/javascript, use capability detection.

Bah Humbug! Javascript is essential to this document template
solution.

That restricts use of the document to environments where you have
absolute control over the client's features, the web is not such an
environment.

Sizing fonts down to of my configured preference results in
unreadable text.

This approach assumes the default browser text is about 16px.

A browser's default text size is an unknown, no assumptions should be
made about it. Leave the default text size alone or specify 100% and it
should work for everyone. If you shrink the default text size down to
62.5% your text will be unreadable for a portion of users.
If your resolution is so high, then you probably adjust the
text size on most sites anyway.
There need not be a relation to monitor resolution, if there is then no
on the fly adjustments are necessary, a simple reconfiguration of the
browser's default size suffices.
At least, my text scales.

Not stuffing up the method to solve a problem you've created is not
smart, many people are unaware of how to enlarge text in their browser.
The proper thing to do is not create the problem in the first place.

There are more browsers than IE and Mozilla, a site published on the www
should be constructed to work on most of them.

The solution is client-based.

The real question is do you have absolute control over the clients.
Given that you seem to be striving for dual client compatibility, and
because you proclaimed this to be a "web development", this does not
seem to apply. If this is for a controlled client environment then this
group is not the correct platform for your query (note the www in the
group's name). Posts to this group are discussed in a www context.

--
Spartanicus

[Robert Neville]

On Mon, 02 May 2005 23:26:50 GMT, Spartanicus
<in*****@invalid.invalid> wrote:

Maybe, I should have included the standard disclaimer. You explain the
whole scenario and people scuff at the unwieldy post; if you are too
concise, you may not provide them with enough information. Generally,
I find a balance between too much and too concise. This approach may
mislead some people to assumptions.

STANDARD DISCLAIMER: My project aims at creating a document template
and navigation approach for my computing notes. These web pages are
for personal usage and will reside on an external hard drive. I will
use these documents to help me remember key computing idiosyncrasies.
These documents number 100+. The CSS style sheet should allow me to
have a consistent format and allow me to update formatting for all
documents at once. The documents should be primarily viewable on
Firefox; Safari, Opera, IE (distant Third) and Netscape (further
distant fourth); and not have any platform specific needs. Thus
server side processing like PHP is not desirable. All processing
should be client based.

[Spartanicus]

Robert Neville <robert_neville310@.no******@hoo.com> wrote:

Maybe, I should have included the standard disclaimer.
That would be unwelcome here since it doesn't say anything about the
specific message. What is part of the posting conventions here is to
quote something of what you are replying to (trimmed down to a minimum).
STANDARD DISCLAIMER: My project aims at creating a document template
and navigation approach for my computing notes. These web pages are
for personal usage and will reside on an external hard drive.
Then they are not web pages and strictly speaking off topic for this
group.
I will
use these documents to help me remember key computing idiosyncrasies.
These documents number 100+. The CSS style sheet should allow me to
have a consistent format and allow me to update formatting for all
documents at once. The documents should be primarily viewable on
Firefox; Safari, Opera, IE (distant Third) and Netscape (further
distant fourth); and not have any platform specific needs. Thus
server side processing like PHP is not desirable. All processing
should be client based.

If these documents are for your use only, why the many different
browsers? If you need to access these documents using clients installed
on systems that are not yours, you then have no control over their
configuration and the requirements are pretty much the same as for web
documents.

--
Spartanicus

[Markus [MSFT]]

"The DOM inspector in Firefox helps,
yet I have not found its equivalent in IE."

There are several solutions out there that provide basic help with
analyzing web content

- DOM Inspector
http://www.download.com/IE-DOM-Inspe...ml?tag=lst-0-1
- DOM SPY
http://www.download.com/DOMSpy/3000-...ml?tag=lst-0-2
- Web Accessibility toolbar
http://www.nils.org.au/ais/web/resources/toolbar/
- IEDocMon: http://www.cheztabor.com/IEDocMon/index.htm

Markus