info-mac@uw-beaver (info-mac) (07/29/84)
From: TIM@MC The Local Editing Protocol paper has some interesting ideas in it, and I think that anyone who is interested in this topic should read it. However, although the basic idea is a very good one, this particular protocol leaves much to be desired. The main problem is that it was designed with rather dumb "intelligent" terminals in mind, and it puts such a great burden on the host that it defeats its own purpose. As far as I can tell, Stallman designed the protocol back in early 1982, hacked up ITS Emacs and the Lisp machine SUPDUP terminal program to test the design, and no one has used it since. Neverthless, other parts of the SUPDUP protocol are definitely worth looking into, (although it was designed primarily for networks). Among other things, it features a flow control protocol that works considerably better than XON/XOFF; a nice graphics protocol, and a "line saving" protocol which allows the host "to tell the local terminal to save a copy of some displayed text, and later refer to the copy to put that text back on the screen without having to transmit it again." [Richard M. Stallman, "The SUPDUP Protocol," MIT AI Memo 644. The source lives in "RMS; SUPAP >" on MIT-MC]. Only a few operating systems support the SUPDUP protocol, but in principle, you could write a simple "SUPDUP Server" for any operating system that supports PTY's. Essentially, it would have to talk SUPDUP to your Mac on one side and emulate a fancy display terminal on the other. Tim McNerney