Yeah I used to code like that a lot. The next step was to work out the "core layout" for the "overlay loader". There wasn't enough memory to keep the whole program resident at once so you had to invoke the loader to bring in the next piece.<p>There was a "front panel" on every computer that showed the registers and memory. You could "single step" through your program to see what was in the registers. When DEC shipped the PDP-11 series the front panel was gone. I didn't know how I was going to debug. (Imagine single-stepping a program that runs on a 3 GHz machine.)<p>Next task was to transfer the program to paper tape using a piece of plastic with a slot for the tape, 8 holes, and a slim rod to punch out the bits. Be sure to leave a couple extra bytes so you can "hot patch" the program. This involved changing the bits into a jump to an address where you had punched new instructions.<p>Oddly though, I don't see the literate programming markup. I guess that's on the next page.
He talks about this here...
<a href="http://www.webofstories.com/play/donald.knuth/27" rel="nofollow">http://www.webofstories.com/play/donald.knuth/27</a>
The thing I love about the handwritten sheets is the small things that aren't perfect. The drift of the commands to the right as they go down the page. The strikethrough of sections.<p>I don't know how to explain it, typically, but notebooks are incredibly intimidating when the examples you see will be perfect end products. Seeing a few of the sheets that made it into the trash are really instructive.
Great information both from the linked document and the fellow posts, thanks to all for sharing.<p>I follow with great interest everything Burroughs related, specially their approaches to safe programming.