Apple gained about one point, but now I think the tide has really turned back the other direction. The economy is helpful. Paying an extra $500 for a computer in this environment — same piece of hardware — paying $500 more to get a logo on it? I think that’s a more challenging proposition for the average person than it used to be.

I wonder what he has to say now that Apple has reported the “Best March Quarter Revenue and Earnings in Apple History” while Microsoft reported their “first revenue drop since going public in 1986″.

Ordinarily I think this issue goes totally unnoticed – there isn’t a lot of info out there on the web about it (at least that’s what I thought at the time – I’ve since found a number of sites covering the topic). Most of the time our buttons have text like Save, Submit, Go – that sort of thing. 

I was flummoxed by this for a while. I was, after all, declaring my padding explicitly, and IE seemed to be rendering it. Rendering it and then some more depending on how much text was in the box. How could I trick IE into behaving like a nice, normal, standardsy browser?

I searched and searched the inter-tubes until I found a simple solution in the comments of a site that, I think, was in Portuguese. Unfortunately I didn’t make a note of where it was so I can’t give credit, nor could I understand anything that was written other than

 overflow: visible;

Now there were other folks on other sites advocating all manner of other, additional style declarations – setting width to auto or to 1% - none of which worked for me. But as soon as I popped that declaration in on my inputs IE’s buttons fell in line.

If you’re experiencing the same issue you might consider adding this to your style sheet.

 * html input { overflow: visible; }

Non-IE browsers will ignore it and, really, you aren’t telling them to do much of anything anyway (inputs don’t generally overflow you know). 

If you find that this doesn’t solve all of your IE button width related issues then I suggest you pursue some further reading.

In our case we have block elements, tables being the main culprit, explicitly set to 100% width inside a container div which will scroll vertically as it needs to. The problem is that IE calculates what “100%” is before it applies the scroll bar. So when the scroll bar is applied IE’s “100%” value is now wider than the actual viewing area. This causes the element set to 100% to be partially covered by the right-hand scroll bar and for a horizontal scroll bar to appear.

After tying to find a solution using conventional CSS or JavaScript to get the offending elements to behave we turned to using IE style expressions. This is essentially what we did.

element_selector {
width: expression(this.parentNode.offsetHeight >
this.parentNode.scrollHeight ? '100%' :
parseInt(this.parentNode.offsetWidth - XX) + 'px');
}

So we’re telling the element that if it is not taller than it’s parent then it should be 100% otherwise it should be the width of its parent minus a specific value (XX): in our case the padding plus the width of the scrollbar.

I’ve successfully applied this technique in a few different places and it hasn’t bothered me since (especially since they are hidden in an IE specific stylesheet behind conditional comments. This isn’t to say I’m a fan of CSS expressions – they shouldn’t be necessary – but in this case they’ve proven to be the simplest and most reliable solution to an annoying IE problem.

In the interest of full disclosure I should mention that I’m no Microsoft fan. I have used Windows for many years and I even have an Xbox 360. So I’m not a full-on hater by any stretch, but if the scale ranges from lover to hater I’m probably closer to the latter.

So, The Mojave Experiment. I think that, from a marketing perspective, Microsoft is missing the mark here in a pretty big way, in a way that is very typically Microsoft. As I see it the tenets of Microsoft marketing are as follows:

1) We (Microsoft) are the best.
2) If you (consumer) disagree with #1 then you are stupid.
3) You don’t want to be stupid do you?

The Mojave Experiment is a great example of this strategy. The assumption of the experiments is that Vista is fantastic and the only reason people don’t agree is that they are too dumb to understand how great it really is.

So they take a bunch of “real” consumers, all of which hate Vista for no apparent reason since none of them have seen it, and demo Vista form them – except they tell them it’s not Vista but a new version called Mojave. Naturally their minds are all blown by Vista’s sheer awesomeness and they all plan to go buy whichever of the 37 versions of Vista is right for them.

Here are, as I see it, the two main problems with this “experiment”.

1) Microsoft is admitting that people hate Vista. Many consumers probably don’t know much of anything about Vista, but now they will know that for some reason basically everyone hates it. Regardless of the fact that everyone ends up a convert in the end people are savvy enough to understand that Microsoft isn’t running this campaign for no reason. It smells of desperation and that generally doesn’t make people excited about a product.

2) Microsoft completely fails to address the reason that people don’t want Vista – their computers can’t run it because it is such a resource hog, it doesn’t work with peripherals – practical things like that. They just show features largely borrowed from OS X and talk about how great they are.

Maybe the Seinfeld campaign will be better. I’m not holding my breath.

UPDATE: Looks like I wasn’t far off about the Seinfeld campaign. I don’t think it’s even two weeks old yet and Microsoft has cancelled it.