|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
Dori Smith: "Neither the Weblog Filter nor the Weblog Favourites have updated in the last two days. I suspect there's a problem with the Weblogs.com XML files." Dori, there was a problem, I think I fixed it. The apps that are reading the XML files should start updating again after the 8PM scan. Craig Burton: "He he he. I love irony." Remember the old maps of the New World. In the 1500s the coastlines weren't well understood and it was a long time before they even knew there was a western coast. In that spirit, here's a hand-drawn map of Microsoft's scripting environment circa July 2001. Now that I have a scanner it's easier for me to do these. Tell me what's wrong or missing in the picture. At the same time, I'm working on a map of what I'm calling The Open Scripting Environment. In that map there's no single rectangle containing the goodness. A different philosophy, a zig to the zag. Thanks to Simon Fell for the pushback on the presence of the JIT compiler. Here's rev 2 of the hand-drawn map eliminating the bytecode interpreter. Note that some old maps of planet earth had the Garden of Eden. Nicholas Petreley: Microsoft's Bait and Switch. "I believe that Miguel de Icaza's enthusiasm for porting the Microsoft .NET development environment to open source as a project called Mono to be naive and potentially dangerous to the open-source movement." I share the same concern and expressed it to Miguel last week. However, if we can also link up diverse scripting environments with SOAP and XML-RPC, there's no reason to worry. Choice is key. Most non-Microsoft scripting languages will never run well inside Microsoft's environment. Make it easy for Perl programmers to participate in SOAP networks without leaving home. Same with Java, Python, Tcl, and everything else. Progress report. Python is on board. Things are looking good with Perl. We've started talking with Sun about baking SOAP and XML-RPC into Java. And there are more behind-the-scenes things that I can't talk about at this time that will alleviate Petreley's fear. Focus on the protocols, that's what's important. As long as we invest in diversity, Microsoft can't control. Instead of a one-party-system, let's have an n-party-system. That's how we guarantee choice, eliminate lock-in, and maintain forward motion. BTW, an interesting detail came out at the open source summit. Dick Hardt of ActiveState reports that Perl does not run well in Microsoft's environment. The problem is that Microsoft's virtual machine is designed to run C-like code, but Perl is not like that. Now I know the solution, we need a DLL-based open scripting architecture, that allows environments to compile and run scripts and have them call back into the environment, much like the architecture we developed on the Mac in the early 90s. Back then it wasn't so interesting because scripting was still pretty small, it was just us and Apple playing with OSA. Ten years later there's been an explosion, and there's another way, beyond XML-RPC, that's needed to integrate. It can be a tough sell to each individual community, as XML-RPC is, because the benefit is that it makes it easy to bridge to other environments. Most communities tend not to see too well outside their borders. But the larger world wants choice. No matter how great your scripting environment, you will eventually meet someone you want to work with who works in a different environment. I posted a subset of these comments on the Slashdot thread for Petreley's article. On this day last year Napster got a reprieve and I was happy!!
|
 
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
© Copyright 1997-2005 Dave Winer. The picture at the top of the page may change from time to time. Previous graphics are archived. Previous/Next |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
