- From: Dave Reynolds <der@hplb.hpl.hp.com>
- Date: Fri, 13 Oct 2006 17:02:21 +0100
- To: RIF <public-rif-wg@w3.org>
I have completed my action: http://www.w3.org/2005/rules/wg/track/actions/146 which was to justify the "limited number of dialects" requirement in UC8. Since most justifications are short I put: [[[ the value of RIF in this use case (reusability of the rules across the enterprise) is only realised if there is a good chance that the dialect Vlad uses is supported widely by the tools other users might have access to ]]] Which might still be too long and could be shortened further. However, I wanted to expand it slightly here ... To me one risk of RIF is that it covers so many different rules languages we risk ending up with practically a dialect for each rule language, which won't achieve much interoperability that way. My interpretation of "limited number of dialects" is that we aim to look for coherent clusters of rule languages. For a supported cluster we aim to have a RIF dialect that all languages in the cluster can reasonably support but which many applications can work within, thus achieving interoperation. If we have two many dialects (or equally dialects which are too rich) then users will still stay locked into the one or few systems which support all the richness of that dialect that they are using. With this interpretation "limited number of dialects" is a candidate for being a requirement that is relevant across the board. I don't think UC8 justifies it particularly more strongly that other UC's I just didn't want it to disappear. Make sense? Dave
Received on Friday, 13 October 2006 16:02:44 UTC