Re: Nice Automator article on O'Reilly
Re: Nice Automator article on O'Reilly
- Subject: Re: Nice Automator article on O'Reilly
- From: has <email@hidden>
- Date: Fri, 6 May 2005 11:51:12 +0100
Timothy Bates wrote:
>After years of waiting, we get yet-another-way-of-doing-things-poorly.
>
>1990ish, we get AppleScript! Yay. It is cool.
>
>For 10 years, basically nothing happens.
Well, nothing shippable anyway. ;)
>2000: We still have Applescript on OS X, but it hasn't changed in a decade
>still has far too few built in functions and no interface elements. People
>talk about AppleScript 2.0
AppleScript 2.0's been vapourware since 1997. More recently rechristened "AppleScript X", with the suggestion it'd be targeted more at professionals and semi-professionals (presumably to rival VB), and still just as vaporous. Frankly I think the opportunity for Apple to do its own proprietary VB-level platform has pretty much sailed, given the time and resources it takes to get something like that to maturity and broad acceptance, and the smart thing to do is to throw their lot in with the OSS languages, particularly Python and PHP, which outside of VB are pretty much owning this part of the field now.
>2003ish: Applescript has not changed much, and instead of Applescript 2 we
>get Applescript studio (Xcode with applescript) which no one uses because it
>has an insane learning curve and the applescript element is irrelevant: just
>write cocoa apps.
Studio's no FaceSpan, but it's amazing what folk are willing to deal with with when the monetary cost is free. Presumably pragmatism beat idealism here: building a tolerable tool off of PB/IB was going to be much cheaper than constructing an ideal one from the ground up. Better to work with whatever limited resources are available and make the best of it that sit and sulk and dream about "If only..." Pragmatic compromises are nothing if not the Mac OS X way.
>2005: Still no changes to Applescript like people want, but instead we get
>automator, which no one will use.
Automator isn't really aimed at AppleScripters, it's aimed at folk who need basic batch processing and common task automating, but really aren't willing and/or able to go near /bin/sh. It's a different niche, and while it remains to be seen how successful it is I think Apple are positioning it very well: as the first step beyond the pure do-everything-yourself-by-point-n-click interaction model of the traditional Mac GUI. Say what you like about DOS and what an improvement Mac OS was over it: this was the one area where Mac OS _completely_ fell down in comparison to DOS. Hopefully Automator will finally fix that final glaring deficiency in the true Mac way.
>Just give us AppleScript with a stripped down interface builder (basically
>the 24U Osax: buy it for $30k for god's sake) and then access the php
>function library (open source) and wrap it all in AppleScript.
>
>Bingo. We have what we want. How hard is that? In a decade?
>
>1. Kill AS Studio
>2. Port the php library into AppleScript
>3. Give an AppleScript interface to unix's commands (like Ed Lai did in
>what, a few weeks? 4 years ago?)
>4. Release a simple webkit editor to allow user interfaces to be made in
>dhtml/css/ using the proven hypercard metaphor (pages = cards, click on
>objects to embed scripts in them).
>
>5. Let everybody know that AppleScript is the way to go, and keep on making
>it better.
A few years ago I'd have agreed this'd make a serviceable solution, but now I don't think it's really worth it: AppleScript itself simply isn't that good a platform to build on. It'd be like trying to build Pages using the original AppleWorks codebase; it's simpler just to start afresh. So, a counterproposal:
1. End-of-line AppleScript. No more new AppleScript language development work; provide only basic maintenance from now on. Like Classic support I can't see it going away any time soon, but like OS 9 there's really nothing to be gained from doing anything more with it. From the users' point of view this will make very little difference, but within Apple it'll release focus and resources to start on the Next Big Thing.
2. Bless Python and PHP as the ways to go for professional, semi-professional and ambitious-amateur development. Ensure both have full and first-rate OSA support. (While OSS folks in both camps are working on this themselves, progress is slow; throwing just a little bit of Apple manpower at this would make a significant difference.)
3. Set out to create a new, truly end-user oriented programming tool that neatly fits the gap between Automator and Python/PHP. Spend the next year watching how Automator does in the field, learning from its successes and failures. Study the state of the art in end-user programming research: programming by example, visual programming, natural programming, etc. and shamelessly crib ideas and experience from current practical applications such as [for visual programming] Alice, Cocoa, Prograph and LabVIEW. Nab a few bodies and set them to work for a bit, trial the results with users to see what works best, rinse and repeat.
4. Profit.
5. Let everybody know that Mac OS X has the best damn solutions at _every_ level of user need, from the slickest and easiest-to-use visual command line that even Aunt Millie will love, through to the badass-est superdynamic flavour of C any l33t hAx0r could ever wish for, and everything else inbetween.
Cheers,
has
--
http://freespace.virgin.net/hamish.sanderson/
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Applescript-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden