Whew. Looks like comments and everything else is working again. Now I wanna tweak Bryan's theme to be XHTML compliant. Any tips?

I'm playing with my themes and trying out Bryan Bell's CSS-based themes. Nice stuff. I love Radio and content-management systems. Don't like how something looks? Click a button and everything changes. You know how hard this is in old-style tools like FrontPage? Geesh.

You heard it here first:

Scoble was wrong.

In 1995 I said Borland would die.

I was wrong.

I had just seen Phillipe Kahn introduce Delphi. He seemed drunk to me and obnoxious. A sore loser, after all, Microsoft had kicked his ass. And he attacked Visual Basic, which was the language that gave me 10 years of pay checks (I was an editor with Fawcette Technical Publications BasicPro Magazine which then became Visual Basic Programmer's Journal).

I still like Visual Basic. It made Windows programming easy enough that a fool like me could do it.

When I saw Delphi I saw something that was technically superior, but that I thought would fail in the marketplace. Much like Beta was a technically superior video tape format to VHS, but failed for mostly marketing reasons.

I was wrong.

My thesis was that Borland was so weakened due to years of poor decisions made by Kahn that it would be irrelevant, at best, and disappear, at worst.

Neither prediction has happened.

Today Borland continues along building technically superior tools to the stuff that Microsoft brings out, but they are starting to get some marketing wins as well.

Which brings me to yesterday. Yesterday I pointed out that TopStyle is an awesome way to build a standards-based Web site.

Little did I know that TopStyle's author Nick Bradbury reads my weblog.

I remembered an conversation I had with Nick in 1995 where I told him that Delphi was the wrong choice and that VB would ultimately prove successful in the marketplace due to Microsoft's market superiority. (Nick had been a VB developer back then, and was just switching over to Borland's Delphi).

I was wrong.

For a few reasons.

1) I thought Microsoft would listen to independent developers and ultimately come out with a product that served their needs. Microsoft didn't. Think that they have? Look at the runtime requirements of .NET.

2) I didn't realize that Microsoft would screw existing VB developers by making VB.NET not compatible with code written for VB 6.0.

3) I didn't realize that the government would get involved in Microsoft's life and slow them down and keep them from rolling over every little company in its path.

4) I didn't realize that marketing guys would take over the VB team and that VB would be developed almost wholly for large companies' concerns.

5) I didn't realize that Microsoft would make an "investment" in Borland (many insiders think this was a payoff to compensate Borland for stealing its leading developers -- in fact, Microsoft's .NET architect, Anders Hejlsberg, was the guy who designed Borland's Delphi and was heavily courted by Microsoft for years).

So, let's look at what Nick's done.

He wrote Homesite (now the standard tool for HTML handcoders the world around). He wrote that application in Borland's Delphi. He then sold that tool to Macromedia. Macromedia now includes HomeSite in most of its Web site development packages.

Nick also wrote TopStyle with Borland's Delphi.

So, why couldn't he do that in VB? My thesis is: the runtime.

The runtime has been VB's advantage and achilles heel since it first saw the light of day in the early 1990s.

In theory, a runtime-based language should be better than one that builds standalone apps.

After all, if you have 30 applications on your system, they can all share the same runtime. These apps would be smaller in total because they could share a common code base (the runtime).

One problem with that: not many people have the runtime, so application developers have to distribute the runtime with each application. If Microsoft had figured out a way to put the runtime into Windows, then this would be an advantage for .NET, but so far Microsoft has not figured out how to do that and due to the way tools are developed (bug fixes come out every six months or so) it'll probably always be impossible to guarantee that any operating system will come with the latest runtime).

The .NET runtimes are about 20MB!

Nick's program itself is about 4MB.

How many sales would Nick lose if he tried to force everyone to download a 20MB file just to run his application? Also, bandwidth costs money. Let's say Nick's app gets popular. All of a sudden he'll have five times the bandwidth costs that a .NET developer will have.

One other thing. Delphi is freaking fast. Try out TopSite. It is fast and responsive no matter what you do (and I'm running it on a 733MHz system here at NEC). Many VB apps I've seen are a lot slower. Yeah, .NET is better in this regard, but Delphi was out five years before .NET, which gave independent developers like Nick a competitive advantage (Nick developed and sold an entire app in that time).

It also gave them the ability to say that Scoble was wrong.

My hat's off to Borland and to guys like Nick who bring us cool new apps to help us do our jobs. Sorry for being wrong. I always figured the better marketing team would kill the better technology. I'm glad to see that wasn't the case in this instance.

Oh yeah, who said there aren't little guys doing cool stuff anymore?