Beaten to the punch

Third, this is going to be a product, not one that I plan to sell (although I may give it to UserLand and they might sell it). It's a new kind of outliner-based Web CMS, that does weblogs and all the other stuff you see popping up here on Scripting. I spent much of the day exploring ideas for how to package, test and then ship this. Should it be a Manila plug-in? Or something that's linked to Manila on a sort of peer basis? Or should it be completely independent of Manila? I want to deliberate on this decision, because I expect to live with it for a long time.

Looks like Dave is going to beat me to the punch on this one. I've been headed in this direction for about a year now. My problem is that I was completely unsatisfied with Radio's outliner and I've had to create a lot of infrastructure before I could even start. I'm at least a month away from serving my first web page, at least, and that's probably pretty conservative.

There's a lot of potential here that's largely untapped by current systems, if you use a properly sophisticated idea of an outline. Combine "transclusion" with "node types" and define a few other things and you've got a very, very flexible writing system.

I've thought about this a lot, and my input into Dave's dilemma would be to make the outline system more independent of Manila, rather then less. You'll probably find yourself replacing almost all of the parts anyhow, depending on how deeply you go into "everything is an outline". You can always transclude a Manila message from outside of Manila, but tying yourself to Manila from the outset may not be a good idea, for the very reason it seems appealing on first blush. The pre-existing Manila system will make it very easy to walk down the paths Manila already walked down, only this time with "outlines".

That may itself be neat, but it is little more. Instead, I suggest going for the gusto and breaking from Manila entirely. It will free your mind and allow the outline system to develop its potential more fully. Manila may be a useful initial platform, but it should be discarded as quickly as reasonably possible.