DigiBarn Systems: Swyft information appliance

The Swyft by Information Applicance Inc.
(1985)

Image and prototype contributed by Jef Raskin

The following image and prototype was generously contributed by Jef Raskin, who was instrumental in its design. This system has possibly never been depicted on the web before. A few more words on SwyftWare and the SwyftCard from Jef next:

At Information Appliance Inc., we decided to use the Apple II, as a development environment. Running our own Forth (tForth), we developed and tested the software until it was our own major tool for every task at hand. These tasks included word processing, spreadsheet, software development, communications (then over phone lines) and data storage and retrieval -- we were strong believers in the maxim that if the product isn't good enough for us, it's not good enough for our customers. Of course, we were aware that the converse was not true. If it was good enough for us, that didn't prove that it was good enough for our customers.

To assure that it was good enough for our customers, we did a lot of testing, and one of the results was that the testees nearly always wanted to know when they could get the software. After you've heard such a request repeatedly, the idea arises that perhaps we had an Apple II software product. First we supplied it as a ROM on a plug-in card, which was called "SwyftCard". This had the advantage of being a piece of uncorruptible and uncrashable software. A lower-cost alternative was then developed, with the software being delivered on a 5 1/4 inch floppy. This was called "SwyftWare".

We sold the rights to market SwyftWare to a company run by a friend in San Diego.

The code was extraordinarily well-documented and there was a test word for every word in the program, as well as a word that ran all the tests as an automatic suite so that we could check for side-effects whenever we made a change. It is to these methods I attribute the nearly unique accomplishment of Information Appliance: a piece of commercial general-purpose software in which no bugs were ever discovered. We had the same splendid results in designing the Canon Cat.

Photos of Swyft prototype donated by Jef Raskin

Prototype with space for LCD screen	Names of people are imprinted on the circuit board
Image84.jpg	ID and serial number (0001)
Image86.jpg	Image87.jpg
Image88.jpg

Prototypes before the Swyft production version
(photos from David Wing)

Serial No. 2 prototype of the Swyft

In David's words:

So, here is Jef's Serial No. 2 prototype as I see it. My example has a non-functioning CRT but Jef told me that the rest of the machine was working when he gave it to me. I think this was a simple FORTH interpreter, and did not carry any of the Swyft "user-application" code in ROM, as the Swyft and Cat did. This was the development series before the 100- or 200-unit pre-production run of the Swyft prototype.

Earlier comment by David:

[this is] the "wrinkled" early Swyft (Forth) development machine, beige case, serial No. 2... When Jef gave it to me around 1989, he noted that the machine worked but for the CRT which was not functional. I also have Swyft prototype No. 10015 (the ur-Cat) which was made in a run of (I think) either 100 or 200 to demonstrate statistical reliability of design and production. It worked fine last time I plugged it in (remember you can't say "turned it on" as it had no power switch, just work, auto-save, and dormant states.

Also see our other photos of the IAI Swyft design prototypes.

The DigiBarn's Swyft card in a modified Apple IIe

Digibarn pages on the Canon Cat

Jef Raskin's homepage

ISDA awards page for the Swyft

Please send site comments to our Webmaster.
Please see our notices about the content of this site and its usage.
(cc) 1998- Digibarn Computer Museum, some rights reserved under this Creative Commons license.