Yes, it was frustrating, but I didn't send off any angry letters or anything. I've been in technical support, and I know the special kind of frustration that occurs when one gets phone calls and e-mails to tell one something that one already knows: namely, that "something" is "wrong".
It's good to be back up!
Well, I figured it out, thanks to a clue provided by Joe Clifford Faust at the Word Foundary.
As you will recall, MacIntosh users browsing this website on Internet Explorer saw my page blown very, very wide. Joe had something similar happen to him, as have others. He eventually figured it out by commenting out portions of his site until something worked and then narrowing down the commented-out code until he found the line that was giving him trouble.
Using this technique myself, I realized that my DIV tag was causing me trouble (again).
You might be asking, at this point, why bother with DIV tags, since TABLE tags do much the same thing and they haven't given me as much trouble. Am I over-enamoured by the fact that DIV tags are fashionable in geek circles and I'm trying to run with the crowd? Well, probably. However, DIV tags do have an advantage: they're more easily controlled by stylesheets, and they are rapidly becoming the standard for laying out web pages. My own workplace is upgrading its standards, and may well drop the TABLE tags for DIV tags. It's good that I learn.
Aha! You thought this website was up just for fun, didn't ya? Well, it is, but it has proven useful for testing various techniques that might come up at work.
Think of DIV tags as boxes. You set up a <DIV> </DIV> code on your HTML, and you've drawn a box. Everything within that box can now be formatted according to a class that you set up on your stylesheet (e.g. <DIV class="myclass">), including font sizes, background colours, or if the box has a border around it. In my case, my web page has three major areas: a header, the main cell (where this post goes) and a navigational sidebar. I tried to set this up as three DIV boxes.
TABLEs are easier to format. Everybody knows what a table looks like: it has a bunch of cells and when you put things into cells, they stay right where they are in relation to each other. DIV tags don't work that way. They're not table cells, they're boxes. They can sit about on your web page willy-nilly. You have to work extra hard to make sure that the boxes line up with each other in the way that you want. Add to this the complication of IE, Mozilla and Opera rendering HTML differently (both on the PC and on a Mac), and you have some users surfing to this website and encountering a mess.
So, I gave up on my all DIV tag layout. To make sure the three key elements of my website stayed roughly where they were in relation to each other, I set up a table, and encountered the difficulties I just experienced.
My problem appears to have been in the CSS variable "float". By giving the main cell a float value of "left", I plunked the cell on the left side of the page, and was able to get the side navigation boxes to run down the right side. I didn't need to do this after I planted these elements inside a table, but I kept the "float" variable on. Well, guess what happened when I went into my stylesheet and removed Maincell's float variable entirely.
So, anybody who finds that their web pages load very, very wide when viewed on IE5 for the MacIntosh, look at your graphics, or check out your CSS stylesheet. Eliminate things like float variables. Play around with the widths. This should help solve your problem.
Anyway, it's an excellent post, thoroughly researched. He did, however, misread me on one point. His comment:
To James, energy is a public good because it benefits everyone -- even non-purchasers -- by improving the economy through cheaper electricity as an input to other products. This is where I disagree with him. You could say the same thing about lumber, concrete blocks, nails, and thousands of other things.
First of all, the analogy between power and building supplies doesn't quite hold true. When a builder runs out of lumber, he can build with metal or composite products. When he runs out of aluminum nails, he could try wood or iron. More importantly, the benefit of the thing being built travels directly to the person who buys it. There is no subsequent benefit derived after the home is bought, and the homeowner shoulders all of the burden willingly.
Power is power. While different sources for power can be found, there is no alternative to its use. You either use it, or you don't (or, you buy yourself a generator). Also, the benefits of power filter out to the wider society, and the burden isn't easily shared. I've already given the example of the benefits derived from a road that aren't easily recovered through tolls, or the benefits derived from running public transit at a loss. Power has a similar problem: do people who travel at night by the light of the streetlamps shoulder any of the burden of providing the power?
Power does sit on the borderline of what constitutes a public good. There is a way of sharing power's burden, by passing on the costs of power in the price of the goods that are made from it. However, I think there is a good argument, in the interests of allowing the freer flow of capitalism elsewhere, to ensure that power is inexpensive and readily available. Even at 4.3 cents per kilowatt hour, Ontario's power is still one quarter of the price paid by average New Yorkers. Even more than our cheap Canadian dollar, this little feature is one reason why Ontario enjoys so much investment.
And I wasn't suggesting that the costs of producing power be subsidized. Until last May, Ontario Hydro rarely subsidized the cost of its power; public ownership meant only that high levels of debt could be achieved to build new powerplants -- something that Ontario Hydro has choked on, but which privatization itself has not helped.
Instead I would suggest that power either be produced, or distributed, at cost (which Ontario Hydro, in many ways, did). I haven't heard that suggested in privatization debates, much. In fact, perhaps a good model would be for the government to retain control over its power distribution network, in the same way that the government retains control over its roads. Government run non-profit power plants, and private power plants could then sell into this market. I think the setup makes more sense than the semi-privatized mess that the Eves government has made (where power is very much subsidized, at taxpayers' expense).
Anyway, just food for thought. I hope that gives you a buzz.