12 ms·

Well, Simula predates PARC by about a decade, so I'm not so sure about AK created OOP. As for Windows and such, sniff around YouTube -- that notion goes back to

by alanlit 10y ago

Well, Simula predates PARC by about a decade, so I'm not so sure about AK created OOP. As for Windows and such, sniff around YouTube -- that notion goes back to the late 50's early 60's. That AK popularized this stuff - sure.

kragen 10y ago
Your history is pretty confused, although it has some elements of truth to it. Let me set you straight. Although there were graphical displays in the 1950s (some of them being simply direct displays of the contents of Williams tubes), GUIs as such only go back to Sketchpad (1961-65, roughly) which sort of had windows, and there were windows and a mouse in NLS by 1968, but without a GUI. Simula (1967) wasn't OO; Alan coined that phrase to describe Smalltalk (1971 at PARC, established 1970, which you'll note is less than a decade after 1967, though begun in 1969), which drew inspiration from Sketchpad, Simula, the B5000, NLS, and some weird magnetic tape format developed by the Air Force where the tape started with a program that interpreted the rest of the tape, among other places. Alan and Ed Cheadle actually did a lot of the development of the ideas that later became OO in the 1966-70 timeframe on a personal computer they were working on called FLEX. You can read Alan's account of this history in "The Early History of Smalltalk", one copy of which is at http://gagne.homedns.org/~tgagne/contrib/EarlyHistoryST.html http://gagne.homedns.org/~tgagne/contrib/EarlyHistoryST.html. Alan and Dan and Adele and other people on the Smalltalk team deserve a lot of credit both for developing the ideas of object-orientation but also for popularizing them, and essentially the modern desktop GUI (mouse, overlapping windows, icons, WYSIWYG, multiple fonts, proportional fonts, buttons, menus, drag-and-drop) was developed at PARC in the 1970s, an effort that included the Smalltalk team but also included a lot of other people working in other languages. Of course, much as I detest Steve Jobs, the really powerful popularizing influences on this stuff were Macintosh and NeXT. (Later, Java.)
mmille10 10y ago
This gets to something that's dogged AK for ages. He created the term "object-oriented," but the way everyone in the field thinks of it is not what he meant. He blames himself for that, using terminology that got people focused on objects. It's not about the objects. It's about messaging and interfaces, and the independence of objects. Objects are just the endpoints for messages. His notions of objects have a lot more in common with how the internet (TCP/IP) works than the way Simula, C++, or Java works. The real abstraction, indeed the real programming, is about interfaces and messages, and their relationships. A short summary of it is to think of objects as servers on the internet, and interfaces as analogous to the protocols used for communication between clients and servers. If you could imagine programming using the structure of the internet, with the same scaling and loosely-bound qualities that the internet possesses, that would be the ultimate in OOP. Simula doesn't get into this at all. It's a class-based language, and what the design really communicates is that it formalizes the notion of abstract data types. C++, Java, and C# do the same thing, in this respect.

by alanlit 10y ago

kragen 10y ago

Your history is pretty confused, although it has some elements of truth to it. Let me set you straight. Although there were graphical displays in the 1950s (some of them being simply direct displays of the contents of Williams tubes), GUIs as such only go back to Sketchpad (1961-65, roughly) which sort of had windows, and there were windows and a mouse in NLS by 1968, but without a GUI. Simula (1967) wasn't OO; Alan coined that phrase to describe Smalltalk (1971 at PARC, established 1970, which you'll note is less than a decade after 1967, though begun in 1969), which drew inspiration from Sketchpad, Simula, the B5000, NLS, and some weird magnetic tape format developed by the Air Force where the tape started with a program that interpreted the rest of the tape, among other places. Alan and Ed Cheadle actually did a lot of the development of the ideas that later became OO in the 1966-70 timeframe on a personal computer they were working on called FLEX. You can read Alan's account of this history in "The Early History of Smalltalk", one copy of which is at http://gagne.homedns.org/~tgagne/contrib/EarlyHistoryST.html http://gagne.homedns.org/~tgagne/contrib/EarlyHistoryST.html. Alan and Dan and Adele and other people on the Smalltalk team deserve a lot of credit both for developing the ideas of object-orientation but also for popularizing them, and essentially the modern desktop GUI (mouse, overlapping windows, icons, WYSIWYG, multiple fonts, proportional fonts, buttons, menus, drag-and-drop) was developed at PARC in the 1970s, an effort that included the Smalltalk team but also included a lot of other people working in other languages. Of course, much as I detest Steve Jobs, the really powerful popularizing influences on this stuff were Macintosh and NeXT. (Later, Java.)

mmille10 10y ago

This gets to something that's dogged AK for ages. He created the term "object-oriented," but the way everyone in the field thinks of it is not what he meant. He blames himself for that, using terminology that got people focused on objects. It's not about the objects. It's about messaging and interfaces, and the independence of objects. Objects are just the endpoints for messages. His notions of objects have a lot more in common with how the internet (TCP/IP) works than the way Simula, C++, or Java works. The real abstraction, indeed the real programming, is about interfaces and messages, and their relationships. A short summary of it is to think of objects as servers on the internet, and interfaces as analogous to the protocols used for communication between clients and servers. If you could imagine programming using the structure of the internet, with the same scaling and loosely-bound qualities that the internet possesses, that would be the ultimate in OOP. Simula doesn't get into this at all. It's a class-based language, and what the design really communicates is that it formalizes the notion of abstract data types. C++, Java, and C# do the same thing, in this respect.