While it is true that HTML5 as a spec is still under active development, certain parts are more mature than others and those parts are the ones being implemented in browsers right now. I've seen examples of both differences being resolved by newer versions of the spec (Google's O3D vs. WebGL for an easy example, although not really a 1:1 correspondence) and differences in fact being created by newer versions (the whole codec debate for example was because the OGG mandate was removed.) So as to whether quirks will pop up when everything is finally said and done, it is tough to say, I'll give you that. But I don't think any quirks will actually impede development, certainly not like how difficult coding around IE6 used to be (even when it was released.)

3. XHTML is still around. 1 and 1.1 won't ever be depracated and browsers will always understand them so there's nothing wrong with that. XHTML lives on in HTML 5 as well, just serve the document as application/xml and close your tags and it's automatically converted. 4. I don't think you've been paying much attention to HTML5 at all from the sounds of it. This isn't really an insult to you but I'd refrain from talking about it until you research things a bit. HTML5 is actually a massive spec and to conform to it you also have to follow the rules on what to do about quirks. Probably the worst thing we've seen from browsers are browser-specific CSS attributes, and those existed long before HTML5/CSS3 and are for mostly trivial things anyways like rounded corners. If you made an HTML5 site using the currently most stable features, it would render exactly the same in any of the current browsers, including IE9. I guarantee you that. 5. Personally I wouldn't code any JavaScript outside of jQuery unless I have to. jQuery is just a really nice, small library that lets you do amazing things, simply. As for tools, the debuggers in Firefox and Chrome are great, although for actual code creation things still seem to be a work in progress. But then again I just use Notepad++.

Well of course it changes, any tech changes eventually unless it's Bad Tech™. But the point is that HTML/JavaScript probably won't ever be thrown under the bus by a single company; in my view having 5 entities work on it is a pro, not a con. It's not a silver bullet but nothing ever is, really. I just see it as the most future-proof tech right now. YMMV.

Well as far as servers are concerned, the NoSQL movement has emerged, and JS is becoming more powerful, so you can always look into that. In fact a whole new stack is emerging thanks to how powerful servers and browsers are becoming that directly challenges the frontend/scripting/server language/datastore setup traditionally used. So if you want new and exciting you can jump headfirst into that. As for desktop programs, those aren't going to change because let's face it, the desktop hasn't changed very much since it was introduced. It looks a bit prettier but the paradigm is still the same as it always was for desktop software. (Thankfully though things are moving to the web more and more which I think is great. :-D But your mileage may vary.)

I can't say I feel very sorry for Silverlight developers right now. They have staked their livelihood on a single company's whims, a company mind you that has a storied history of changing its mind whenever it pleases. "Oh wow, Win32! Time to-" "Err, .NET is in now? Well OK, let's-" "Oh, WPF? I guess that makes since with Vista-" "Silverlight? Really?!" Just to give some examples of how spastic their desktop development toolchain has been. You may as well buck up and learn HTML/JavaScript because that stuff is proven, platform/company-agnostic, and is only going to grow in the future. Now I'm also a bit of an open web app zealot but I wasn't always that way. I still use Windows and I've always loved C#'s syntax and features - I just wouldn't stake my job on them given the choice.