Archive of UserLand's first discussion group, started October 5, 1998.

Re: Programming ideas and time to mature

Author:Dave Winer
Posted:8/30/1999; 6:12:23 AM
Topic:Programming ideas and time to mature
Msg #:10216 (In response to 10209)
Prev/Next:10215 / 10217

Interesting question! I wondered the same thing as I read about Algol-60. There was Algol-68 and then C, which was Algol-72, and Pascal (75?) and then Modula and Oberon and Java, and..?

But scripting goes back to the 70s not the early 90s. In my opinion, Unix was and is a scripting environment. And the early IBM PC was a scripting environment too, with no less than two scripting languages burned into ROM -- the command line language and Bill G's BASIC.

I remember when I first learned the Mac OS at a programmer level in 1986, looking thru Inside Mac for the scripting language, it had to be there, I even found the early Script Manager but was totally disappointed to see that it was for internationalization and had nothing to do with running scripts.

I suspect that scripting goes back even before Unix, but that's where I first encountered it. My feeling about the Mac was that it only solved half the problem, but did that half very well. The interactive interface was ground-breaking. But there was no way to automate the OS.

Someone is going to chime in about HyperCard if I don't say something. Yes yes HyperCard was scripting, but it didn't integrate with the OS or the apps. So it was an either/or -- either you could have scripting and no apps, or apps and no scripting.

Now object databases are a whole other matter. As far as I know there was no precedent for what we did with Frontier in the late 80s. I wasn't copying anyone in the design of our ODB. Frankly, it wouldn't have happened if Apple had been willing to bundle our early scripting language with the Mac OS. The odb was in that early version, but it was invisible. It probably would have stayed invisible, and the Mac would have had system-level scripting about five or six years before they eventually did have it.

So maybe that's a hint as to why adoption takes so long. Corporate stuff gets in the way. NIH. Politics. Personalities.

Hey, I sent an email to John Ousterhout a few weeks back pointing him to XML-RPC and asking for his support. He sent back a thank you note, said he'd take a look. Want to take bets on how long it takes for Tcl to embrace XML-RPC at an official level? What about other scripting languages like Python and Perl? Do you think that Java ever will embrace it? And when they do, will it look anything like ours? And this is just one slice in time.

And why has it taken so long for PhotoShop to be scriptable? Certainly there's no good technical reason for it taking so long. It's been eight years since System 7 shipped and ten years since it's been technically possible.

I think there's the answer to your question. Fear is what keeps the rate of adoption of good ideas going at a slow pace.

There are responses to this message:

This page was archived on 6/13/2001; 4:52:15 PM.

© Copyright 1998-2001 UserLand Software, Inc.