- From: Sankar Virdhagriswaran <sv@hunchuen.crystaliz.com>
- Date: Thu, 8 Feb 2001 14:46:57 -0500
- To: <ietf-dav-versioning@w3.org>
Lisa, > (1) Core useful, options useless. > (2) Core useful alone, options useful. > (3) Core useless, core + options useful. > (4) All of DeltaV useless. > This is one way to phrase the argument. Your argument discretizes the choices where there is a continuum. Therefore, I would like to interject a different conceptual view point. As I understand things, the spec. goes from serial joint authoring support to more and more parallel joint authoring to merge support. For those interested in supporting features that involve parallel joint authoring, semantic consistency of the specification as it moves from serial authoring to parallel authoring is important so that whatever marketing decides, the engineers can digest one model of consistency management and implement features as time permits. The most recent note from John Vasta on this thread sort of points this out. For that crowd (I am one of them), the fact that the specification is the way it is helps us identify and deal with semantic consistency issues particularly when it comes to implementation. Wading through 4 different documents is not fun. For some one like you who has been involved in the discussions intimately, parceling the pieces does not raise an issue due to familiarity with the issues and how they were resolved. For some one like me who is not tracking all the issues, who is hoping to read the spec. in detail once it is done, I don't have any context. In my mind, the tradeoff is between losing context and therefore consistency/ease of implementation vs.the one-large-document digesting issue. Since I will be spending more time in implementation than in reading and digesting a spec. document, I vote for optimizing the larger value. Sankar
Received on Thursday, 8 February 2001 14:46:05 UTC