It's fun, and possibly even profitable, to speculate on what technologies are going to be "main-stream" popular in the future. By main stream I mean:

• When you mention a technology to other developers and 90% of the time they've discovered it already.
• When you see the jobs relating to the technology appearing on popular job boards.
• When all the big guns players have started investing in that technology in some way.

Do you every find yourself learning a technology, and wandering if it will be a good choice? Personally I think I've made some good choices over the last 7 years. I'm no innovator, but perhaps a fairly early adopter...

• I was reading about TDD and Refactoring, and using NUnit in .NET back in late 2000 when it was first released. Before that I was making my own TDD tools in Visual Basic 6.0! Obviously MS have started taking Unit Testing quite seriously now, but I don't recall seeing them giving us tools back in 1999. I think the SmallTalk and Java folk had been using some form of TDD for years, but this wasn't talked about much in the .NET world.
• I was watching Ruby On Rails from 2004, and 1 year later in April 2005 I had started offering it to clients and had produced my first commercial rails solution. I think Microsoft will copy the lessons of success in RoR as much as possible over then next 3 years.
• I was using NHibernate s beta releases back in 2003. That project has matured greatly over the last 4 years, and it's still on the up. Microsoft look like they'll release their own ORM tools soon (they've promised them for a while!), so it looks like ORM was a good technology to back.

What do I think is hotting up now ready for mainstream?

• Business Intelligence (BI). Businesses collect so much data that they can't use. Data loading, cleaning, predictive analysis etc are going to hit main-stream within 5 years I think. It will mean the difference between business success and failure. They will be selling software next to the USB memory sticks in Dixons for £s;9.99 that says things like "Clean up your enterprise data and predict your customers buying patterns in 5 minutes"
• Distributed Computing. I think that in 10 years we'll all be hitting F5 in visual studio, and our apps will automatically run against any spare CPU cycles on local trusted machines. Clustering on databases will be soooooo easy, like clicking a check box during installation that says "Allow MSSQL to automatically outsource work to other network machines". lol. This will probalby mean further growth in Space Based Architecture, Grid computing, distributed memory systems etc (and Linux!).
• Open Source. Open Source seems to have been booming for a years now. I think this will continue, these days the quality of open source software in almost any category is astounding.
• Virtualisation: This has been hotting up for years, but I think it's just beginning to break into consumer computing. Soon we'll be taking our virtual PCs on holiday on USB sticks, and then loading them up in internet cafes. They'll be like the most portable computer in the world, that can be backed up in seconds and sent to other people by email. We'll each have 20 virtual machines for different purposes - media jukebox, work, games. What's better, well be able to run them all side by side on the same hardware.