|
Note 39
Harry Forsdick, one of the originators of the Diamond multi-media mail system, recalls “During the Christmas break time in 1984 I decided to take the Diamond components (text, graphics, images, …) and see if I could recast them into a different application – one that took advantage of the real-time communication capabilities of the local area network. The idea was to allow people at two workstations to collaborate over the editing of a single document. This turned out to be a rousing success and caught the attention of a lot of people at BBN as well as at our sponsors. I wrote about this in Explorations in Real-time Multimedia Conferencing, a paper which it turns out has been cited as prior art in several patent infringement defenses.
When I came back from presenting this paper, Terry Crowley, the lead developer at BBN in multimedia applications, ripped apart the rapid prototype programming efforts and over the next year or so made an infrastructure that came to be known as MMConf. The system that Terry developed intercepted input events (keyboard key transitions, mouse movements) and output events (operations that changed the display) and distributed them to other conference participants. A floor control mechanism based on tokens was used to resolve simultaneous activity and various policies about how to use this mechanism (“raise hand, be recognized,” “interrupt at will, take the floor,” etc.). We described this system in MMConf: An Infrastructure for Building Shared Multimedia Applications. Subsequent to this work, a lot of people regained interest in the topic of real-time collaboration including many of our competitors in the government contractor community. Of course, Douglas Engelbart had demonstrated many of these ideas with the now famous NLS demonstration at the Fall Joint Computer Conference in 1968 and an associated paper written in 1975, NLS Teleconferencing Features.
When I came back from presenting this paper, Terry Crowley, the lead developer at BBN in multimedia applications, ripped apart the rapid prototype programming efforts and over the next year or so made an infrastructure that came to be known as MMConf. The system that Terry developed intercepted input events (keyboard key transitions, mouse movements) and output events (operations that changed the display) and distributed them to other conference participants. A floor control mechanism based on tokens was used to resolve simultaneous activity and various policies about how to use this mechanism (“raise hand, be recognized,” “interrupt at will, take the floor,” etc.). We described this system in MMConf: An Infrastructure for Building Shared Multimedia Applications. Subsequent to this work, a lot of people regained interest in the topic of real-time collaboration including many of our competitors in the government contractor community. Of course, Douglas Engelbart had demonstrated many of these ideas with the now famous NLS demonstration at the Fall Joint Computer Conference in 1968 and an associated paper written in 1975, NLS Teleconferencing Features.
Copyright 2012-2022 @ xbbn.org under a Creative Commons license.
