The Alto itself wasn’t more user-programmable than a modern PC. It booted into a very arcane monitor program and required a few typed commands to load an OS from the disk. One of those “OS”s was Smalltalk, which was user-programmable, but you can run similarly-programmable environments on a modern computer with a lot less trouble!
Yes but, as I understand it, it was expected that many end-users would operate the Alto using Smalltalk. That is, a significant use-case was a user-programmable OS. That is, it was expected that people using the Alto for work would do so in the context of a user-programmable environment.
Even programmable environments nowadays are, for the most part, just a small subset of the functionality in a given OS, and are almost all targeted towards specialist programmers, not end-users.
That’s a fair point. I’m skeptical about user-programmable systems, though. Automobiles also started as user-tinkerable, but it turned out most people didn’t want to take apart and reassemble cars, they just wanted to drive them. But those who were interested took shop classes and could work on cars in their garages. Same with computers.
(Tinkering with Smalltalk was fun but could really get you in trouble. I made some extensions to pop-up menus once, but several times a bug meant I completely broke pop-ups in the whole system, and it turned out to be impossible to evaluate code or back out my change without popping up a menu and creating another recursive level of debuggers. The only recourse was to abort and relaunch from the last saved image.)
I strongly disagree - look at the sheer number of non-developers using VBA to improve their lives. User-programmability is extremely popular whenever it’s available, even in trivial ways. AutoCAD, Emacs, Excel, Word, game modding …
Probably not for the office worker of the future - those were using Neptune and Bravo, and that experience was refined into what became the Star/ViewPoint.
I’d argue while Alto was hugely important for its vision of the office of the future, Ethernet, etc., in terms of the GUI, it had a lot less influence than its commercial product successor, the Star. Except for Smalltalk, GUIs on the Alto tended to be ad-hoc experiments, whereas the Star had not just a recognizable desktop metaphor and a common toolkit, but concepts taken for granted in offices like roaming profiles.
Yeah, even by 1980 Smalltalk still had a pretty primitive GUI. Partly this was because the only graphics primitive it had was bitblit, which copied/filled rectangles. So for example it couldn’t clip to non-rectangular regions, which meant you couldn’t draw into a partly-obscured window. In practice, drawing only occurred in the front/active window.
You also couldn’t just drag windows to move or resize them — you had to right-click and choose Move or Grow from a menu first. And don’t even get me started on the weirdo scroll bars.
The Alto was a first-generation machine, and the second and third generations were more powerful and capable. The Dolphin is what became the Star hardware, and there were a bunch of OSs on it like Tajo, Cedar, and InterLISP that were in many ways more advanced than ST80.
Well there is some truth to this, there’s also a lot that we have that they didn’t think about. Things like 64 cores on a single chip, machines we can talk to, and a worldwide web which while it does have a lot of bad information it also has a lot of good information. And finally the one browser is now most of my app server. And that sense it’s completely different from that model.
“Some” truth? The article never claimed the Alto had everything we have today. But it had embryonic forms of a lot of it. (FYI, I have used an Alto, and other Xerox systems, and spent three summers interning at Xerox in the 80s, so I’m familiar with this stuff.)
If you look at the mind-map style chart at the end of the article, it shows the Alto’s influence on a lot of what you mentioned.
Hardware research at PARC led to VLSI and ever more powerful CPUs. Alan Kay and the others bet that future computers would be vastly cheaper/smaller/faster, so they designed systems that were hugely impractical at the time but would be practical in the future we now live in.
PARC invented Ethernet, which made networking practical, and Ethernet protocols that were influential in the design of TCP/IP.
Notecards, built in InterLISP on Altos, was one of the first hypermedia environments and a major influence on HyperCard and the WWW.
A lot of distributed computing and client/server technologies were pioneered at PARC — mail servers, directory servers, file servers.
Apps in your browser depend on JavaScript, which is a direct grandchild of Smalltalk. The DOM APIs are based on the MVC design invented in Smalltalk. And text editing in your browser and editor is directly descended from the Bravo editor.
I wish we could live more in the Alto’s world. Machines designed to be user-programmable general-purpose computers would be great.
The Alto itself wasn’t more user-programmable than a modern PC. It booted into a very arcane monitor program and required a few typed commands to load an OS from the disk. One of those “OS”s was Smalltalk, which was user-programmable, but you can run similarly-programmable environments on a modern computer with a lot less trouble!
Yes but, as I understand it, it was expected that many end-users would operate the Alto using Smalltalk. That is, a significant use-case was a user-programmable OS. That is, it was expected that people using the Alto for work would do so in the context of a user-programmable environment.
Even programmable environments nowadays are, for the most part, just a small subset of the functionality in a given OS, and are almost all targeted towards specialist programmers, not end-users.
That’s a fair point. I’m skeptical about user-programmable systems, though. Automobiles also started as user-tinkerable, but it turned out most people didn’t want to take apart and reassemble cars, they just wanted to drive them. But those who were interested took shop classes and could work on cars in their garages. Same with computers.
(Tinkering with Smalltalk was fun but could really get you in trouble. I made some extensions to pop-up menus once, but several times a bug meant I completely broke pop-ups in the whole system, and it turned out to be impossible to evaluate code or back out my change without popping up a menu and creating another recursive level of debuggers. The only recourse was to abort and relaunch from the last saved image.)
I strongly disagree - look at the sheer number of non-developers using VBA to improve their lives. User-programmability is extremely popular whenever it’s available, even in trivial ways. AutoCAD, Emacs, Excel, Word, game modding …
Probably not for the office worker of the future - those were using Neptune and Bravo, and that experience was refined into what became the Star/ViewPoint.
I’d argue while Alto was hugely important for its vision of the office of the future, Ethernet, etc., in terms of the GUI, it had a lot less influence than its commercial product successor, the Star. Except for Smalltalk, GUIs on the Alto tended to be ad-hoc experiments, whereas the Star had not just a recognizable desktop metaphor and a common toolkit, but concepts taken for granted in offices like roaming profiles.
Yeah, even by 1980 Smalltalk still had a pretty primitive GUI. Partly this was because the only graphics primitive it had was bitblit, which copied/filled rectangles. So for example it couldn’t clip to non-rectangular regions, which meant you couldn’t draw into a partly-obscured window. In practice, drawing only occurred in the front/active window.
You also couldn’t just drag windows to move or resize them — you had to right-click and choose Move or Grow from a menu first. And don’t even get me started on the weirdo scroll bars.
The Alto was a first-generation machine, and the second and third generations were more powerful and capable. The Dolphin is what became the Star hardware, and there were a bunch of OSs on it like Tajo, Cedar, and InterLISP that were in many ways more advanced than ST80.
Well there is some truth to this, there’s also a lot that we have that they didn’t think about. Things like 64 cores on a single chip, machines we can talk to, and a worldwide web which while it does have a lot of bad information it also has a lot of good information. And finally the one browser is now most of my app server. And that sense it’s completely different from that model.
“Some” truth? The article never claimed the Alto had everything we have today. But it had embryonic forms of a lot of it. (FYI, I have used an Alto, and other Xerox systems, and spent three summers interning at Xerox in the 80s, so I’m familiar with this stuff.)
If you look at the mind-map style chart at the end of the article, it shows the Alto’s influence on a lot of what you mentioned.