SiriScript
Today I faced a long list of alarms on my iPhone, and decided that I wanted to clean them out. The typical iOS “Edit” interface puts a red “delete” button next to each item, and upon tapping it you must then confirm it by tapping the explicit word “delete” at the other end of the item. Suffice to say: for a list of any significant size, this is very tedious.
On a whim, I decided to give Siri a shot at simplifying the process. I long-pressed the home button, and uttered: “delete all my alarms.”
Too bad there’s no equivalent for clearing out all your photos. You still have to empty the camera roll from the Image Capture application on the Mac.
Something like Siri on OS X could radically simplify everything from the most basic commands to much, much more complex tasks.
2 Comments RSS · Twitter
> Perhaps unsurprisingly, my current side project is to create an "AppleScript killer" - a scripting language that supports whitespace within names and uses punctuation much as it's used in written English.
> No guarantees it'll ever get anywhere, natch, but still - if fifteen years of learning, using, and explaining AppleScript isn't a powerful motivator to put the boot in, I don't know what is. :)
Hey, you announce it on Michael's comments section, you can't back down. That's the rule here, right? :D
In any case, that sounds great and I am really curious to see where it goes.
The spoken aspect is a bit of a red herring: written (typed) text is also a perfectly good UI, and probably more appropriate to OS X. The hard challenge is in taking an ad-hoc string of mostly arbitrary words and turning those into a precise machine operation. Funny enough, the "AppleScript" system - specifically the Apple Event Object Model - isn't an entirely awful match (much better than OO-style IPC) since it's already query based, but there's still a big, big gap between what a user will casually express and what an AEOM interface will actually accept. (And forget Automator: that thing's just an evolutionary dead-end where end-user programming is concerned. Ironically, text is the future; just not the way programmers write it.)
At the same time, I think there's a lot could be done to reduce the gap just at the scripting language level: Papert's Logo was highly learnable and composable, but its potential was never adequately explored (as a disruptive technology, it went down with the orthodoxy like a lead balloon). That HyperTalk/AppleScript took its basic design from Pascal, not Logo, will always be its single biggest mistake, since it made syntax and semantics insanely complex from the start. All the subsequent nonsense of trying to make AppleScript look simple instead of throwing out and redesigning it to be simple (as Logo is/was) sprouts from that.
Perhaps unsurprisingly, my current side project is to create an "AppleScript killer" - a scripting language that supports whitespace within names and uses punctuation much as it's used in written English. (My current language, kiwi, achieves the former but still falls short on the latter.) Still at the design stage, so I don't even have a working prototype yet (although kiwi has explored some of the issues already). But I've been getting up to speed on Swift with a possible view to having a crack. No guarantees it'll ever get anywhere, natch, but still - if fifteen years of learning, using, and explaining AppleScript isn't a powerful motivator to put the boot in, I don't know what is. :)