Posts by MidiMagic

Joeprogrammer

What is in your code box is one line of "browser-proofed" html, followed by plain text.

It should render as source html code, not as an actual html object.

For security reasons, you should not have control over the user's refresh, forward, and back buttons.

I just realized the utter impossibility of doing this and making it fair.

You do not have control over the following:
- The browser the student has.
- The Internet Service Provider the student subscribes to.
- Whether the ISP burdens users with popups in exchange for free or low-cost service.
- Security settings.
- Firewalls.
- Traffic loading on the ISP.

So I can see some problems:

- The page stops loading in the middle, but the timer starts.

- The page loads, but JavaScript is turned off.

- The page loads except for the script.

You need to use a form, with an input tag of type "button", along with JavaScript, to do this.

Here you go:

http://www.w3.org/TR/xhtml1/

Scroll down to part 4.

Actually, you should not use html attributes as variable names.

Part of the problem is that war is against the religious beliefs of some liberals. They will go to any length to be rid of it.

But Presidenbt Bush hasn't committed any crimes, so he can't be impeached.

On the other hand, President Clinton not only committed crimes, but admitted that he committed them on TV. But he dismissed them as irrelevant.

He was not charged with sexual activites, but with lying in court (perjury) about them.

The only reason Clinton wasn't convicted is that Senate Democrats valued party unity more than justice.

I just had my code (in the proper code pseudotags) in a post show up in the advanced editor as evaluated html, showing the form items I posted code for. This was about 5 minutes ago.

This didn't happen while I was editing, but when I hit the back button after I finished the edit and saw the result (I was stepping back to the main topic menu).

I have a theory:

The text is somehow being evaluated twice. The first time, the & codes are changed into html. The second time, the html is evaluated and rendered.

I just got a FF upgrade just before this started.

Here is the relevant code.

The JavaScript uses this simple loop to put the values in the form boxes. It works fine, except for the delays.

for(cs = 1; cs <= 3; cs++){
    tmps = 'cm' + cs;
    document.forms.prgrss[tmps].value = saband[cs];
    erx = errif[cs];
    document.forms.prgrss[tmps].style.backgroundColor=eco[erx];
  };

After this code runs, the JavaScript enters a long series of nested loops which have nothing directly to do with the form items shown. The results are written into other similar text boxes.

Here are the form items. They are inside a form tag pair named "prgrss", along with other form items.

<input type="text" id="cm1" name="cm1" value=" " size="32" />
  <input type="text" id="cm2" name="cm2" value=" " size="32" />
  <input type="text" id="cm3" name="cm3" value=" " size="32" />

The code itself works. The ONLY problem is that the display does not update until after the long loops finish:

- IE leaves the old values and colors in the text boxes until the script ends. When the script ends, the new values appear first, and the new colors appear about a second later.

- FF replaces the values at the time the script writes them, but leaves the old colors in the boxes until the script ends. When the script ends, the new colors appear.

I copied this from another thread because it became buried. So far, all I got were kludges that changed other parts of the page to fix the problem.

Divs don't always work exactly the way you want them to, especially when they are inside other layout tags. I gave up after SIX hours of trying to get a simple three-column set of divs
to line up correctly inside a pair of li tags. It always either dropped one div to a lower area, or took some of the text below the ending li tag and put it on the right of the three divs, or failed to display correctly on one browser while working on the others.

Here is code that fails. Somebody please tell me why. You will note that the set of divs inside the li pair fails, but the ones down below the ul pair work correctly. And IE messes it up differently than FF does. The ONLY difference between the two is that one set is inside the ul list's li pair.

And giving up the bulleted list is NOT an option. The rest of the real page looks stupid without the bullets.

<html>
<head>
<style>

.x3col {float: left; margin: none; border: none; padding: none; width: 33%;}

</style>
</head>
<body>

<ul>
  <li>
    <p>Here is the list:</p>
    <div class="x3col">
      <ol>
        <li>one</li>
        <li>two</li>
        <li>three</li>
      </ol>
    </div>
    <div class="x3col">
      <ol>
        <li>four</li>
        <li>five</li>
        <li>six</li>
        <li>seven</li>
      </ol>
    </div>
    <div class="x3col">
      <ol>
        <li>eight</li>
        <li>nine</li>
        <li>ten</li>
        <li>eleven</li>
      </ol> …

The problem is that people were using tables to create page margins, borders, and other objects which can't be readily done without style sheets.

There is absolutely no reason to not use tables for tabular data, or for forms with rows and columns of data entry.

The place where the division is hazy is the creation of columns. W3C wants us to use multiple div boxes instead of tables. The problem is that it doesn't always work.

Div works fine for making columns on the top level, but it falls apart if the columns are inside some other object (such as being part of one item in an ordered or bulleted list), or if the browser window is narrower than the contents of the columns.

Since table is not deprecated, I suggest that you use it if the div method falls apart and you don't want it to. Maybe someday W3C will come up with a design that does not fall apart.

Also, if older browsers that don't do styles absolutely must show margins, the page will still have to have the table in it to produce the margins.

It does matter in Firefox, which follows the W3C standard. Firefox does not recognize uppercase or mixed case versions of styles and attributes in doctypes that prohibit them.

The style "height" must not have a capital "H". The "h" must be lowercase.

height = OK
HEIGHT = wrong
Height = wrong

The use of uppercase letters in tags, syles, and attributes is deprecated by the W3C standards. They will fail when used with future browser versions. Get used to it now, so you don't suddenly have to do a ton of work on your pages later.

Most browsers on UNIX servers already fail when given the wrong case. UNIX has no automatic case conversion.

IE is not following the W3C standard, and wrongly makes these case-insensitive.

There is a Ford Windstar in my driveway, and I couldn't ask for anything better, unless someone builds the flying car in my avatar.

The properties that are invalid are *all* related to the display of the banner advertisements which, unfortunately, just don't render the same in IE and FireFox without specifying the additional parameters, as IE and Firefox appear to have different default behaviour when it comes to the rendering of IFrames. I probably could replace them with CSS and will most likely do that later today. Thanks for the catch.

unfortunately, I have discovered that many advertising gimmicks provided by third-party sources are not XHTML compatible.

Caught in the act!

After this delay happened, the ad space to the right of the Quick Reply window showed a message instead of the ad:

"The server is not responding."

One more thing. The "type" attribute should be "text/css", not "text/CSS".

You didn't get an answer because you didn't tell us what it was doing wrong.

Note that all CSS attributes are lowercase. The way you spelled 'Height' is not allowed. IE wrongly allows it.

Are you talking an anchor (a) link, or a value in the element?

It's the order in which your items are listed in the html. Your right list is immediately after the title banner.

You have to put the stuff you want on the left in BEFORE the stuff which must be on its right.

If you are dynamically adding the stuff which goes on the left after the page loads, you need to reserve space for it in the html. Otherwise, it doesn't render correctly until you scroll down to the bottom and then scroll up again.

How can I get JavaScript to display values and attributes when they are written to input boxes by the script, instead of waiting for a rather lengthy (10 sec) calculation to finish?

- IE does not change the boxes at all, until after the script finishes.

- FF displays the text when it is written, but won't display changes in attributes (such as background color) until the script ends.

But both display the changes immeditely when an alert box pops up.

In my application, I want a "wrong value" error condition to show up immediately, so the user knows this before the script ends (so he can visually check values while the script is still making other values). But the alert box interrupts the processing.

In IE, the error message won't show up until after the script ends. The old values and error conditions stay on the screen for the duration. So the user doesn't know to check until the script ends.

In FF, the error message changes when the error appears or disappears, but the yellow error background does not appear or disappear until the script ends. So the user won't notice it until the script ends.

Does anyone have a trick that is not browser-dependent?

I don't think you can do that, unless you can store the elapsed time in a cookie.

Where are you planning to save the values to complete it with? A cookie is the only place you can save anything on the user's computer.

Use CSS.

The height tag is deprecated (due to be discontinued), but the CSS height attribute is not. height: 100%; works in most cases. It does not work in two cases:

- Where you are trying to make the item expand to just fill the browser window (which is always impossible if the page is to work on all computers).

- Where there is no box object to take a height attribute.

In your case, there IS a containing box object (the td tag of the outer table), so a 100% height attribute in a CSS class used in the table tag should work. Unfortunately, not only does it not work, but IE and FF do it different ways.

I have another suggestion: Use a single table.

You can use the colspan and rowspan attributes of the td tag to make stuff line up, like this:

<style type="text/css">
.wfl {width: 100%;}
.pink {background-color: #ffaaaa;}
.blue {background-color: #aaccff;}
.puce {background-color: #eeaaf8;}
.gray {background-color: #cccccc;}
</style>

....

<table class="wfl gray">
  <tr>
    <td colspan="3">Bigtitle 1</td>
    <td colspan="3">Bigtitle 2</td>
    <td colspan="3">Bigtitle 3</td>
  </tr>
  <tr>
    <td class="blue">Big 1 little A</td>
    <td class="blue">Big 1 little B</td>
    <td class="blue">Big 1 little C</td>
    <td class="puce">Big 2 little A</td>
    <td class="puce">Big 2 little B</td>
    <td class="puce">Big 2 little C</td>
    <td rowspan="3" class="pink">Big 3 little A</td>
    <td rowspan="3" class="pink">Big 3 little B</td>
    <td rowspan="3" class="pink">Big 3 little C</td>
  </tr>
  <tr>
    <td rowspan="2" class="blue">Big 1 little D</td>
    <td rowspan="2" class="blue">Big 1 little E</td>
    <td …

You are never going to have a screen-filling design for all users, for the simple reason that different computers are set up for different screen settings.

For example, my father still uses 800 X 600 because of his bad vision.

Negative pixels?

Not until you master speling.

And we won't do your homework for you.

Caveat Emailor.

I don't have Outlook installed on my computer, so there is no way you could somehow get my computer to use it. I have only the proprietary emailer my ISP uses.

Every user's email must go through an email system he is actually subscribed to, or through your server-side database page made to fake an email send.

You certainly do not have the right to change my email settings or cause me to use a toll system I don't belong to.

Frankly, I am always suspicious when someone wants to make a custom email page instead of just printing the email address on the web page. The thought of malicious software connected to the send button always comes to mind. With an email address and my own emailor, I know it's safe.

innerHTML is IE-proprietary. It is NOT in the W3C DOM.

Some versions of FF can use it, others can't.

For the last time, Midi, you don't get down off of elephants. Elephants don't have feathers, down, pinion, or otherwise.

Sorry, I missed the first one. Did you edit it into the post while I was reading the unedited version?

On the VII = I, try moving the left leg of the V between the two Is:

II-I = I

Now, can you anagram this phrase so it makes another phrase which is equal to its value?

"eleven plus two"

You are right about the car keys and the duck.

Use string processing to take away the I.
IXX - I = XX

What's round on both ends and high in the middle.

How many grooves are on the average phonograph record?

heart.

:icon_mrgreen: Elves and dwarves.

I want ironclad independently verified scientific proof before anybody does anything about it. So far, all I have seen is bad science.

1. Life as we know it is a dirty trick.

2. Using the product voids the warranty.

3. Every time someone perfects a foolproof device, someone else perfects a better fool.

4. As soon as you tell someone the truth, somebody else changes what the truth is. Example:

A to B: "Your box is on the kitchen table."

C enters room: "Hey! Whose box was that which I just moved from the kitchen table to the closet."

5. Nothing will make a hidden defect in a digital product show up faster than releasing the product for sale will.

6. A CD player is just as useless without CDs as the CDs are useless without a CD player.

7. Once you standardize your music into one recording format, that format will be discontinued.

8. Actions speak louder, but politicians STILL won't shut up.

9. If you set n mousetraps, the (n+1)th mouse gets all of the cheese.

10. Nothing can make a job take longer than imposing a deadline does.

11. A late bus is no excuse to give to an early boss.

12. Never underestimate the power of stupidity.

13. You can tell when Democrats and Liberals have no valid argument in any debate. That's when they start using pejoratives, calling names, twisting the words of others, and all of the other Politically Incorrect things they continually scold others for doing.

When either bad code or code that is undefined (in the Doctype declared) is served to a browser, an error condition occurs. Different browsers handle these errors in idfferent ways.

Oddly enough, the more compliant the browser is, the more messed-up the page looks when an error occurs.

I guess it depends on your browser and ISP.

Several points:

Other sites I use just have a note above the post window, reminding users to use quote or code tags. The watermark is hard to read, and is annoying.

Other sites have buttons with the actual tags they make on the buttons. This site has undecipherable bbeilroglyphics.

The ones from Word are known, but the following ones are strange:
- AA (it's blurry, so that means Alcoholics Anonymous)
- A/A
- the world with an infinity sign
- A crossed-out infinity sign
- A cartoon talk balloon
- A number sign (or hexadecimal???)
- A TV Set
- What looks like a tax form (eeyutch!)

Might I suggest some icons with equally ambiguous meaning?
- Spongebob Squarepants
- A clown throwing a pie
- A tuba
- The three balls of a pawnshop (meaning that two-to-one you won't get it back)
- The half of the world Microsoft left off the IE busy sign.
- A brick
- A muddy tennis shoe
- A record changer

And put the code button on the quick reply window.

It's already here. As I type, I am watching a "Mega-Disasters" on History channel predicting exactly that.

Also, don't forget "The Day After Tomorrow".

IT will cause more pollution, as MS and Intel keep making people upgrade their computers every 3 years.

And don't forget all of the NTSC TV set solid waste when HDTV becomes mandatory (thanks to Bill Gates - we had a compatible system until he stuck his nose into it).

You will never "find" time for anything.
If you want time, you must make it.

But daylight-saving time makes you too sleepy to use the extra time.

Do not follow where the path may lead.
Go instead where there is no path and leave a trail.

Git off'n my propitty with that trail, or I'll pepper ya with my scattergun!

:icon_mrgreen:

I wish there was a way to send the spammers a spam sandwich.

When it gets stuck, I usually don't have a refresh to hit. And if I am posting and there is a refresh, it erases my post.

Using deprecated elements in XHTML will cause unpredictable borwser behavior.

I always check my pages with W3C before deploying them.

I can't stand the first 3, because I can't read text which is on top of graphics.

I like the ones with a plain font and the graphics separated from the text.

There are fundamental differences between the way IE displays box objects and the way Firefox displays them.

Firefox obeys the standards, and places the margin, border, and padding OUTSIDE the defined size of a box object.

On the other hand, IE crams hem all inside the box object's defined size.

The trick is to not use margin, border, or padding on a box object you plan to define the size of. Instead, nest another box object inside or outside the one with the defined size, so you can put your margin, border, and padding where you want them.

I always HATE it when some clown of a webmaster decides I shouldn't have my carefully orchestrated toolbars. So I add that site to my "Don't go there" list.

I hate it even more when they somehow change my toolbar.

You can't do anything to a page to change its Google ranking, except to provide valid meta search strings. As long as the URL is not changed, the Google rank won't be changed by what you do.

I discovered that the links to a page have nothing to do witrh how Google indexes it. I have one page on my site which was designed for me to bypass downloading my entire homepage when I start my browser. I am supposed to be the only person who knows it is there, because nothing but my browser home links to it. But I found it on Google the other day.

Note that your Google rank will change if someone else makes a better page. That's the nature of a ranking (which is why it is a bad statistical measure).

There are some things that will KILL your
Google rating:

- Repeating the same string more than once in your meta list (e.g. "football, football, football").

- Having things in your meta list which have nothing to do with your page (e.g. "pro football" on a page about fish).

- Dynamically added content (e.g. in JavaScript) will probably not be seen by the bots. The bots don't click buttons.

If you want content to line up with links, don't use frames. Use tables.

Check for a typo in earlier code. Here are a few causes of grief that can vary with different browsers:

- Forgetting to close a quote on an earlier tag.
- Closing " with '
- Not using the & codes for certain characters.
- Javascript improperly changing code.

It is best to not change the colors of these items, sinc most browsers can block your styles on links.

Check security settings in IE7.