There are some ideas that never quite die - like "Man's destiny is in the stars" (1940s), "This is the year of Unix" (1975), "CASE tools will eliminate coding" (1980) and so on.

The latest one is the idea that WebApps can and will replace desktop apps. The truth is, though, that to accomplish this goal, you end up grafting so much stuff into the browser that it ends up being a mini-OS all unto itself.

In the end, that really just provides a mechanism to deliver software on a per-use basis (and hopefully, extract some per-use fees in the process). However, to get there, we have to do a couple of things that I think will always remain an impediment to this concept.

To start with, you need fast, reliable, always-on, always-available Internet. And while we're there - remember, now you're squirting a lot more data over that pipe, so it better be cheap too. It's interesting that the people who promote WebApps aren't the people paying for the pipe...

Then there's the question of security. I don't even trust someone to use my laptop - why in God's name would I then turn around and let some company I don't know, or have any say in, store my sensitive data?

Finally, there's interoperability. With apps, I get an actual file I keep on my system. If the app becomes popular, other app makers will try to be compatible with that file. If I decide I like their apps better, I can migrate. But with WebApps, my data is up on their site in God knows what format.

WebApps have a definite place and it's an important one - but the simple reality is - desktop apps aren't going anywhere anytime soon and will always be the mainstay of personal computing. How they work and how they get delivered may change, but it'll still be a desktop of some kind running an application locally.

Click to read the article this is in response to.