Date: Fri, 12 May 2000 17:45:47 -0700 (PDT)
| From: | Joe Williams |
|
altintdev@webtv.net Reply-To: unrev-II@egroups.com |
| To: | unrev-II@egroups.com |
| Subject: | Current Proposals |
Hi,
When a document is brought into the DKR it is translated to DKR format if necessary. The translation is only necessary if the original format is not XML based - .doc or .txt, for instance. After that operation, the original is saved for future reference, but it is 'dead' that is changes to it cannot be tracked by the DKR except by human intervention.
An example might be where an original documenet (A) contains some fact, like the current estimated speed of light in a vacuum at 0G. That A document is imported to the DKR and translated to produce B for use by other participants. Perhaps later, that fact gets cut and pasted into a new DKT/OHS information item C.
Now, suppose the author of A issues a rev to A updating the speed of light. This cannot directly replace A. When it is brought into the DKR, all that can be done is to make it A' and alert B that the underlying doc has changed, leaving it up to author of B to find the changed fact and put it into B. making it B'. When B changes, if the specific fact was identified appropriately, it can be promulgated to C, making it C' automatically. The point is that the A material, while is may be a part of the DKR, is not under control of the DKR. The user must be aware that once A is brought into the DKR, the users must encouraged to forget A and focus on evolving B,
The collaborative environment has to start somewhere, and from the point of view of the DKR processing, it must start with B. Now there surely will be some way to create a flag with A' that alerts B, and in turn C, that A has changed to A' so the users can be aware that some change has been made to A, but I can't see how B can be alerted that any specific content has changed in A. That would be up to the author of B to discover.
Likewise, at the output, suppose a document C' is published from the DKR then revised outside the DKR to state a new number for the speed of light (unofficially C''). Obviously, there is no way for the DKR to 'know' that the document has been revised until it is reinput to the DKR. Now should B' and C' be automatically updated? I guess if the info item was tagged as originating in A' and used in B' and C', and the tag has been retained in the external edit, then we may want to automatically update B' and C', but certainly we would not want to go in and try to revise A' automatically.
If this process seems confusing at this point, it may be because I don't get it, but it is more likely that the real news is that it only gets more difficult from here. It is the core problem we face in this augmented knowledge management system.
Thanks and Best Regards,
Joe
Joe Williams
altintdev@webtv.net