nrh@inmet.UUCP (09/25/84)
Has anybody put remote cancellation into "notes"? It's (so far) the one great disadvantage notes has with respect to news. It seems like it wouldn't be all that hard (famous last words).
essick@uiucdcsb.UUCP (10/03/84)
Why haven't I put remote cancellation into notes? Here are a few reasons. @ Once the data arrives on my machine, it's my property and I don't want some random Joe on the other side of the network to zip those babies away. Particularly if it's something I want to forward to my archive machine. @ remote cancellation leads to the following scenario, which is also why notes doesn't let you delete a note or response if there are later notes in the discussion: 1. person A writes a note 2. person B writes a well-thought-out, rational, coherent response. 3. person C writes a really stupid response 4. person D writes a response that says "the previous response is really stupid". 5. person C goes back and deletes his response 6. Person D now looks like a bozo for (apparently) saying that person B's response was really stupid. Steps 5 & 6 are avoided by not letting you delete a note/response if there are later responses in that string. The situation gets more complex if you let people do remote cancellation. @ Since the notes networking is passive in the sense that it doesn't happen the instant you write an article but later when a daemon comes around to package all the new articles (which looks a lot like what batching, doesn't it!) the user typically has a reasonable window to decide that he didn't really like what he wrote and he can edit the text. All of a sudden, you don't need remote cancellation, the reason you wanted to cancel the article has been taken care of in a purely local manner. Anyway, those are some of the reasons that I didn't implement remote cancellation of notes. It opens a number of problems with consistency, you can get people screaming about the information on their machine belonging to them, and the way I implemented notesfile networking you almost always have a chance to correct it before it gets sent out over the network. -- Ray Essick, University of Illinois -- ihnp4!uiucdcs!essick essick%uiuc.csnet@csnet-relay.arpa