I talked yesterday about the history of the Xerox PARC Alto machine, which is a computer from the 1970s that is still influencing your smartphone today. On December 10th, forty years to the day after the presentation of the Alto to Xerox management at Xerox Future Day, the Computer History Museum hosted a retrospective on the Alto, featuring two 40-year-old machines and half a dozen of the forty-years-older engineers who created it and the software that ran on it. Forty Years Later Forty years after that Xerox Future Day, 10th December this year, the Computer History Museum organized an event around the Alto. There were two Altos there, and several of the original engineers who developed the hardware or software. As David Brock said, introducing the evening: It's 40-year-old hardware, running 40-year-old software... and the experimenters are 40 years older. What could go wrong? Indeed, before the evening even got started, the backup Alto had suffered an unrecoverable disk error and so there was only one (which, happily, performed without problem all evening). One thing David Brock emphasized is that the Alto is very limited. A modern desktop computer is 1,000X to 10,000X more powerful. So all the software on the Alto (and all the demos that evening) were running on 1% of 1% of what we have today, not just in processor performance, but in memory and disk too. One thing I hadn't realized was that the initial programming was all done in BCPL, which is a precursor of C developed by Martin Richards, who was one of my undergraduate professors. Almost all programming in the Cambridge Computer Laboratory in that era was done in BCPL (including most of my assignments!). Bravo The first part of the demo was Tom Malloy and Charles Simonyi showing Bravo. This was the first editor that worked the way something like Microsoft Word does today. It supported multiple proportionally spaced fonts, it kept the screen up-to-date as editing progressed, and it created the now-familiar cut/copy/paste model for moving text around. Happily, the laser printer was invented at PARC at almost the same time, so could create documents of typeset quality. Another benefit of Bravo was how easy it was to learn to use, as compared to earlier editors. Indeed, the spouses of many of the researchers came in and used it, creating Ph.D. theses, PTA newsletters, and so forth. Organizations outside of PARC suddenly were blindsided by the technology. When the wife of one of the researchers submitted her Stanford Ph.D. thesis in two copies as required, the librarian insisted on knowing which was the original and which was the copy. Of course, she had simply printed the document twice so the concept didn't exist. They did many demos for VIPs and prospective customers; one was to an executive from a Japanese company. One of their party tricks to show the accuracy of the system, was to laser-print out a document onto a transparent viewgraph (remember them?) and then hold it up against the screen to show that it was identical. "Ah," said the Japanese executive, "What you see is what you get." That was soon abbreviated to WYSIWYG and the phrase stuck. Icarus There was a lot of influential work done on VLSI Design at PARC too. Lynn Conway (see The Book That Changed Everything ) was there, as was Doug Fairbairn, who was the fourth employee at VLSI Technology and hired me and relocated me to the US. He wasn't at the evening event, but he had provided a video of Icarus, which was the layout editor that ran on the Alto. Of course, being only black and white, seeing just what was going on was a bit of a challenge. Luckily, in that era, there was only a single layer of metal. Markup and Draw Next up was Bob Sproull showing Markup and Draw, which were the first editors for graphics. He used the Alto, including Draw, to create a camera-ready copy of "Newman and Sproull", for many years the standard textbook for computer graphics courses, probably the first book to be delivered by the author in camera-ready laser-printed form. Draw is an Adobe Illustrator-type program that operates on lines, arcs, text, shading. Markup was a paint type program that operated on actual pixels. To change brushes, you used one of the mouse buttons, and a menu of brushes appeared. This was the first popup menu. Later systems such as Smalltalk, made extensive use of popup menus, by then looking much as they do today. Laurel Next Doug Brotz showed us Laurel (it became traditional to name products at PARC after plants found in the Sunset Western Gardening Book). Doug said that a number of people had launched the project, including Roger Needham. When I was an undergraduate, he taught me everything I know about operating systems. He went on to become the head of the department, and later, the head of Microsoft Research in the UK. The Alto didn't just have a bitmapped screen. It had a network interface, disks, and lived in an environment with file-servers and interfaces to the outside world. So it was the perfect setup to develop an email system. You only need one glance at Laurel running on the Alto to realize it is email. It looks very similar to modern email user-interfaces. PARC didn't invent email, but until then it had only been used by programmers to get technical work done. Laurel made it accessible to lawyers and executives, and eventually everyone. John Warnock (of Adobe and Postscript fame) suggested adding a graphical language so that it would be possible to send images in Laurel, and that happened. Email was no longer just simple text, and closer to what they are today. Smalltalk Another project at PARC was Alan Kay's and Adele Goldberg's (and others) Smalltalk. There was a demo of that, showing just how far into the guts of the Alto you could program on the fly. Plus some video of middle schoolers learning how to program using Smalltalk. Many of the features of modern object oriented languages were pioneered in Smalltalk (and other languages developed at PARC, such as the later Cedar programming language). Demos One message of the evening was how important demos are, compared to just reading things on paper (especially academic papers). It takes a live demo to show what these capabilities can really do. 12/1968 the "Mother of All Demos" 1972 Article in Rolling Stone Magazine about using a mouse 11/1977 Xerox Future Day 1979 Demos to Steve Jobs and other Apple engineers What Didn't They Invent? In Butler Lampson's 2006 retrospective, he pointed out that using an Alto was very similar to using a modern computer. The user interface with overlapping windows, menus, icons, a mouse, cut/paste and more were all there. Things that were missing were: Spreadsheets The Internet; the Alto was networked, and Arpanet existed and was used for long distance file-transfer, but nothing like the Internet we have today Integration; each tool being independently developed and they didn't always play nicely together Speed; the Alto was 0.3 MIPS (your cell phone is about 10,000 times faster) Xerox made billions on the PARC technology in laser-printers, which also required other people to have taken some of the other ideas and run with them so that desktop publishing came into existence. The diagram below shows where the Alto (and its successor, the Xerox Star) fit into the ecosystem of products in 2006. That was before the iPhone introduction, but if you look at your smartphone today, you can still see shadows of the ideas developed in the Alto and its software. Sign up for Sunday Brunch, the weekly Breakfast Bytes email.![]()
↧