On Wednesday, July 02, 2003 4:42 PM, Jony Rosenne <[EMAIL PROTECTED]> wrote: > I cannot agree with some of these statements. My comments are > inserted. > Jony > > Only invalid for Modern Hebrew. > > No - it is true also for Biblical Hebrew and any other. The extra > vowel belongs to another letter, which is known to exist but isn't > printed.
Not printed but also not coded! That's the issue. You don't have any code point assigned that means: "implicit Hebrew letter"; I'm curious to know which letter(s) is/are implied: if there are only a few, then may be they could be assigned separate codepoints for each semantic value). As long as this extra *implicit letter* will not be coded, there will remain a problem with the NF forms. Some other scripts have such "implicit" letters encoded, or "fillers". Not Hebrew... Using a CGJ (or a new codepoint to define) to code this implicit letter is not a bad idea, and it does not require changing any combining class value.
