Netscape Followup
Interesting email today after Oh Netscape!.One point was made loud and clear, even if releasing the source code might not work for Netscape, it does change the rules about how the Internet works.
I have no argument with that. But I do offer a caveat. Let's focus on interfaces, not source code integration.
An example
We clearly need something like the Open Scripting Architecture that's cross-platform and not Microsoft-owned. Personally, I don't care if it's a Microsoft spec, but many other people do. That means that the various scripting system developers should sit down and work it out. Emphatically, don't integrate Perl with Navigator at a source code level. Do the math. Multiply the number of scripting languages by the number of possible host environments, and you'll see why we can move forward more quickly if we agree to interfaces.
Navigator bloated vastly over the years. I think it would be a great goal to set to have Navigator slim down. One way to do that is to have really high performance well-designed interfaces, unintegrate source.
Evangelism
It's time to think interfaces, whatever economic model you choose to develop for.
I've been evangelizing XML-over-HTTP communication, and every day I believe more strongly that that's the way it's going to go.
Navigator could be a leader in this area. Maybe I've been talking to the wrong person by emailing with Marc Andreessen. Maybe it should be done in public, broadcast to the world? OK, let's give it a whirl.
If there is a Navigator developer community, is anyone interested in hooking it up as a client and server using messages that are compatible with the ones outlined on our XML site?
Want a juicy idea? Make the client side work like the scheme we're implementing in Frontier 5.0.2.
Send me mail and I'll post it on the Mail page.
Dave Winer
|
This page was last built on 4/6/98; 4:26:46 PM by Dave Winer. dave@scripting.com |