Close. By the way there is a bk _scat command (sccs cat, not poop) that dumps the ascii file format so you can try this and see.

The delete needs to be an envelope around the insert so you get

  ^AD 3
  ^AI 1
  this is the first line in the first version.
  ^AE 1
  ^AE 3
  ^AI 2
  this is the line that was added in the second version
  ^AE 2

That whole weave thing is really cool. The only person outside of BK land that got it was Braam Cohen in Codeville, I think he had a weave.

It's sort of suprising then that a delete doesn't just and end-version on the insert instead:

  ^AI 1..2
  this is the first line in the first version.
  ^AE 1..2
  ^AI 2
  this is the line that was added in the second version
  ^AE 2

This way the reconstruction process wouldn't need to track blocks-within-blocks.

Interesting. "^AI Spec" where Spec feeds into a predicate f(Spec, Version) to control printing a particular Version? Looks like you could drop the ^AE lines.

Sounds like equivalent representations no? Limit the scope of the I lines or wrap them in D lines.

Probably missing something. Both are working on one file at a time and have some form of changeset. One is adding from back to forward (kind of) another is from forward to back (rcs). Not sure where the reduction of work coming from.

SUN must like the scat names :) I used to use scat tool for debugging core files.