From: Peter Constable (petercon@microsoft.com)
Date: Mon Mar 30 2009 - 09:53:12 CST
From: André Szabolcs Szelp [a.sz.szelp@gmail.com]
> Bakonyi's proposal's fault is very simple to explain.
>
> It only encodes a small subset of historic signs which he feels to be
> "the true script".
That doesn't go against any Unicode / UCS principles; it simply reflects a disagreement within the expert community as to what is the inventory that needs to be encoded.
Suppose two parties disagree on the inventory, A saying that a text element x needs to be encoded, while B saying it should not. If B's claim is that x can be represented by some sequence of characters there is consensus to encode, then that's something that needs to be resolved. But if B's claim is that x doesn't merit representation while A has a clear use requirement, then the committees may disregard B's objections.
This is, of course, simplifying: if the situation is not clear enough to the committees, then they might accept a subset, waiting for clarification or consensus before accepting x, or they might decide to hold off on everything.
> Bakonyi's proposal is -- in general -- not necessary to be considered,
> as it is a true subset of both alternative proposals (as it has been
> demonstrated by N3532: Mapping between Old Hungarian proposals in
> N3531, N3527, and N3526 by Michael Everson.
Well, that may depend on why Bakonyi excludes some characters. (I haven't reviewed all these docs yet, so don't know the reason.)
> The question of Gábor Hosszú's proposal is slightly less trivial, but
> it can be summed up saying that it counters Unicode principles by
> encoding a large number of precomposed ligatures wich cannot be
> analyzed as characters in their own right.
I assume you mean that they *could* be represented as sequences of characters which there is consensus to encode. If so, then it does sound like that would contradict one of the encoding principles.
Peter
This archive was generated by hypermail 2.1.5 : Mon Mar 30 2009 - 09:56:51 CST