4 ms·
Logo was created by Seymour Papert. He formed his own company in 1980, called Logo Computer Systems Inc., and they produced versions of Logo for various home c
by thought_alarm 8y ago
Logo was created by Seymour Papert. He formed his own company in 1980, called Logo Computer Systems Inc., and they produced versions of Logo for various home computers in the early 80s, including the Apple II.
http://el.media.mit.edu/logo-foundation/what_is_logo/history.html http://el.media.mit.edu/logo-foundation/what_is_logo/history...
Apparently they're still around: http://www.microworlds.com/company/profile.html http://www.microworlds.com/company/profile.html
The version I used in 1983 was for the Coleco Adam computer, which uses the same video and sound chips as the TI99/4A. This version of Logo supported 32 sprites, 16 color graphics, 4 channel sound, and had a built-in sprite editor.
The language itself was quite slow, but it was very easy to come up with all sorts of interesting animations using the 32 sprites. You give a sprite a speed and direction and off it goes. It was a lot of fun and I spent many hours with it as a kid.
- reaperducer 8y agoFound it! Byte magazine, August 1981, page 32. The letter was from Stephen again, identified only as “MIT Logo Laboratory.” ”Texas Instruments is a very generous friend of MIT, and some people here are embarrassed about a situation that might threaten one of its sales programs. So, in an effort to appease its corporate friend, MIT has been stalling the distribution of Apple Logo.” The whole thing is available on archive.org.
- effish 8y agoI don't know that there was any stalling, as it was licensed and released by Terrapin in 1982. This person then did work for Logo Computer Systems on the competing version of Apple II Logo. I'll message him and ask him to respond.
- effish 8y agoAhh, looking down below at the first response there's a link to a '95 comp.lang.logo post with a great looong item by Leigh Klotz on the history of that. His claim was that it was legal wrangling between NSF and MIT on the licensing that caused the delay. This seems highly believable.