I meant the collation order as it applies to Hebrew, of course.OK, so you mean not the collation algorithm but the default (DUCET) collation data for Hebrew. There are still question marks about the adequacy of this data, and there is a clear need for tailoring to meet the needs of biblical Hebrew users if the perceived inadequacies are not addressed within DUCET. The amended or tailored data may need to distinguish between different orders of combining marks e.g. to ensure that a spelling check for Yerushala(y)im is not matched by the invalid form Yerushali(y)am. This case is not in fact an example of a sequence of combining marks in the same class, but it would become such a case according to some of the proposals Philippe and I have discussed. Currently the only such sequences in Hebrew are sequences of accents and so of significance for collation only at the lowest level; that is a consequence of the allocation, now generally considered misguided, of unique combining classes to the Hebrew vowels.
Jony
Or are you in fact trying to make an argument that these unique combining classes are appropriate? I can see the beginnings of an argument for the principle of unique classes, but one that makes sense only if those classes are chosen according to the logical and typing order.
-- Peter Kirk [EMAIL PROTECTED] (personal) [EMAIL PROTECTED] (work) http://www.qaya.org/
