Specifically the Mac issue was in prior version of Solpart - which at the time was advertised as Mac compliant and didn't render on any Mac browsers. I agree this issue has improved substantially. But at the time it was a deal breaker for almost all of my clients that had to be overcome.

Jon - I think you were the one who posted regarding the css issue on the ASPNet forums. I wanted to address some of the points you brought up.

If I remember right - you stated you couldn't justify the trouble of doing a css layout when you can basically bang one out in tables (without hacks, etc..) in much less time.

I agree with your frustration with css layouts being difficult. What I've found (and it's rather humbling) is that I've had to go re-learn html and css from the ground up. As I've come through that process I've found that once you know the browser differences and a few (minimal) hacks, it's really quite easy. Almost as easy as html.

It was almost more than I cared to bother with to learn css layout techniques. My clients have been the driving force. They simply have demanded it. With Zeldman's css evism movement taking hold - it's going to become more and more an issue.

What I did find (much to my suprise) was an amazing speed difference in css based designs. Even partially css driven DNN skins seem to render much faster on the page - leading to greater client satisfaction. This is important when you're toting around 70k of Solpart's javascript on every page (and postback). I realize compression helps - but it's not always possible with certain modules.

Regardless of css driven design or not - what we're after (or I'm after) is valid xhtml (which keeps getting confused with css driven design). People can keep using table driven skins all day long - but I feel the core should generate valid xhtml markup that passes w3c standards. That's what my clients (more very day) are seriously demanding.

Firefox on Mac didn't do any good when the vast majority of the folks on Macs (that I spoke with anyway) used Safari and then IE at the time this was a problem. To be honest - I didn't even check Firefox on Mac. I don't even know that Firefox was much of an issue at the time like it is now.

You have a point on the .js being cached. However - it still impacts initial page loads. There's a lot to be said for that. My point of mentioning it is that smaller = better and faster. There are significant advantages to css driven designs in these areas.

The other part of css speed is the browsers themselves. They (if my understanding is right) don't render a table until it has all of the table and contents down. With css the broweser starts rendering earlier with what it has.

Regardless - I'm getting stuck in to the css design argument and that's still not my main point. Let Zeldman fight that battle. My point is still that DNN should generate valid html markup - that's seperate from the whole css driven design.  

If I'm exaggerating - it's not intentional. I mean to have meaningful discourse. )