so I was able to use a primitive magnetic encoding sequencer to read robot programs from a flat magnetic tape. We had some antique (literally 25 years old) diskettes in the computer history museum, and I fed one to the robot, and then "saved" all the programs written so far to the "diskette" then, i used a mechanical diskette-reader for a desktop computer (which needed lubrication on its rusty motors and slides and quite a bit of fiddling to work) to read this magnetic tape "diskette" the filesytem format is nothing recognized by 'mount' so i was forced to simply dd the binary of the diskette's image to a file. I have posted the file here: http://spaz.org/~jake/robot/robofloppy.2 and then i edited that with vim to extract the programs within. here is a simple program that makes 7 poses, and repeats: http://spaz.org/~jake/robot/baby.txt this is a more raw version of all the jobs i found on the disk image, including some fragmentation. I don't understand the filesystem format, so it's just luck that so many of the files were intact in one piece. http://spaz.org/~jake/robot/jobs.txt as you can see, a program is a list of coordinates (the six hardware axes of the robot arm) and then a program referring to those coordinates. it can be more complicated than that of course, for example when we start using base or robot coordinates for 3d printing, but here it is anyway. this stuff is documented in this PDF: http://spaz.org/~jake/robot/479236-17-Communications.pdf and also, i think we might find the serial communications format used by the "diskette drive" to communicate with the robot, so we can skip the magnetic media entirely and send programs directly to the robot over serial. please let me know if you're interested in this stuff and want to try something. -jake