There isn't any harm in doing so, but I'm not sure how valuable it is, either. In my "firewalled" section of the requirements, I address the need for private vs. public DKRs. You "publish" from the private DKR by putting into the public version, I think. That implies a way to keep track of *which* public DKRs it has gone into, so that only version-differences are sent when publishing subsequently.

Along those lines, an email forum is a "diary" of interactions. The reduction step consists of abstracting the thoughts they contain, effectively replacing them in a subsequent version of the document.

The diary system at Rod's site is difficult to follow because (given the technology available at the time) everything is done with links.

What you want, though, is for much of the material to be "inline" instead of linked. When you publish a "reduced" version, the material that was originally inline becomes relegated to a link -- a link which is identified as something that points to "supporting information" or "original arguments" or the like, so you know you don't want to follow it unless you are particulary interested in deeper information.

Where things get messy is when I do one reduction and you do another. We now have competing versions of the summary. The original discussion now has a higher-level fork, where my document points to the original version, and your document points to the original version, but both of these are under some *other* node -- where they may themselves be summarized, or where evaluations by different folks may lead to one of them being chosen as the "better" reformulation of the issues.

Understanding what happens at such times is one of the things we need to tackle in the use-case analysis.

Sincerely,

Eric Armstrong
eric.armstrong@eng.sun.com