> From: Peter Kirk [mailto:[EMAIL PROTECTED]
> Well, this does not deal with the scenario which I had in mind, and > clearly presented some time ago, in which users are searching the > Internet, or some private but extensive collection of texts, for a > particular word or phrase, in Hebrew or for that matter Moabite etc or > even Phoenician. Currently such a search would need to match Hebrew > characters and also a variety of Latin transliterations. (Hopefully over > time the use of Latin transliterations will fade, or at least become > more standardised as transliterators can use real Unicode characters > with diacritics and not ad hoc ASCII-based solutions.) But if Phoenician > is separately encoded, and at least some palaeo-Hebrew, Moabite etc > texts are represented with the Phoenician characters, searchers will > need to search for an additional encoding. For that matter, searchers > for texts written with Phoenician glyphs will also be inconvenienced > because some such texts will be represented by Hebrew characters. In > such a case the user cannot convert all texts to Hebrew characters in > advance, the folding must be applied by the search engine. > > Is this a realistic scenario? Is it one which really requires folding > together of Hebrew and Phoenician? What does anyone else think? Sure, that's a realistic scenario. But it takes about 1 extra second to type abc OR def rather than just abc as the search criteria, and that achieves the desired result. I don't think the use of Latin transliterations will fade all that quickly. Peter Constable
