Ha, well Chinese is quite a bit more sophisticated than a stop sign or a hospital sign—and it’s also quite a bit more difficult to learn. In that analogy of language to UI, Chinese would be like Cubase or something that does in fact build logical, complex systems out of graphical components. It’s also quite a bit to learn how to use.
And Hugh, I think a graphical way of doing stuff would be fantastic! The great thing is that Tinderbox already has a wonderful way of describing complex systems: it’s own map! If attributes and temporary states could be expressed as nodes on a map, with operators as links, there could be a graphical way of setting up “the wiring”. It wouldn’t be like in EagleFiler where you select from 10 things and choose from 4 things and then type in “hospital”, but it would actually map to a reasonable degree, Tb’s actual capabilities. Like Cubase though, it wouldn’t be something you could learn overnight. I’m not opposed to some way of doing things graphically—but would be strongly opposed to it as a replacement.
Whether that is something a single developer could tackle, I don’t know, it seems like a lot to me. Describing Tinderbox itself, visually, not just the information you put into it, would be almost as big a job as coding Tinderbox, it would seem to me.
I think Tinderbox does have a broader appeal than Maya, though, too. It to a degree has a level of specialisation that is only found in programs like Maya, but I wouldn’t quite put it to that degree. So I agree with you on that point, that it appeals to a wide audience. Lots of us have notes, and lots of us desire to do more with those notes, yes, and there aren’t many programs in between Tb’s all-in approach and a sea of programs that are barely more capable than Finder+Spotlight for that.