The dilema noted by Clive Featherstone of a completely satisfactory technical solution versus layering on a still-not-staisfactory solution is, unfortunately, increasingly common in many standardization area particluarly those where there is a substantial sunk investment and user base in legacy systems. I would propose that one approaches the problem by (1) working on the development of a "completely satisfactory technical solution" for ALL the required functionalities and then, and in the context of (1) (2) develop interim solutions, migration strategies/paths, etc. generally as well as required for particular user bases (e.g. the MS-DOS) community In short separate (a) the totality of the needed/desirable from the (b) immediately do-able, need to recognize the reality of sunk investments. This having been recognized, mapped out and accepted, the only remaining contentious question then becomes one of time lines (pardon the pun). Trust that this is of some use - Jake Knoppers