From: Karl Pentzlin (karl-pentzlin@acssoft.de)
Date: Thu Oct 30 2008 - 17:42:08 CST
Am Mittwoch, 29. Oktober 2008 um 22:02 schrieb Michael Everson:
ME> On 29 Oct 2008, at 20:15, Karl Pentzlin wrote:
>> However, I am a little bit concerned about the fact that the Szekler-
>> Hungarian Rovas proposal scatters over 4 separate BMP blocks, two of
>> them being one-column blocks.
ME> I assure you that isn't going to happen. The proposer does not really
ME> understand the UCS. Please see
ME> http://www.dkuug.dk/jtc1/sc2/wg2/docs/n3531.pdf
ME> for a proposal more in line with our encoding practice.
Regarding the Szekler-Hungarian Rovas proposal
http://www.dkuug.dk/jtc1/sc2/wg2/docs/n3527.pdf
I presume you refer to the Historical Ligatures block proposed for the
SMP when you say that the proposer does not fully understand UCS,
as there is the argument on p.21 that these ligatures have to be
encoded as compatibility characters as they are encoded in a font
from 1995.
Is this assumption correct?
Are there other points where you see a misunderstanding of the UCS?
Besides the historical ligatures, the proposal appears to me somewhat
more comprehensive than N3531, as it relates more to the very recent
activities of the user community regarding the character set, and
takes more into account that the script may be used boustrophedon
directionality (i.e. changing right-to-left and left-to-right every
line), by including the punctuation marks within its script block.
Regarding the Historical Ligatures: it is not clear to me (maybe I
have overlooked somewhat in the text of N3527) to which type they
belong:
like Latin U+00E6 ae: Things which look loke ligatures but are in fact
separate characters, as the ligated form is contrastive in its use to
the unligated form.
If so, then they have to be encoded.
like Latin U+FB01 fi: Ligatures which are the result of regular font
processing when the constituents occur in sequence, without carrying
any meaning and without need for a user's choice except in special
cases for which ZWNJ is appropriate (like in German Fraktur writing).
If so, they have not to be encoded and even do not need any mention
in the context of Unicode/UCS.
like Latin U+FB06 st with bow: Ligatures which are to be selectable
by the user for aesthetical reasons, without carrying any meaning and
without harm when they are not supported by a specific font.
If so, they do not need to be encoded, and may be advised to be
generated using ZWJ or things like OpenType mechanisms.
Do the historical ligatures proposed in N3527 consistently in one of
this groups?
- Karl Pentzlin
This archive was generated by hypermail 2.1.5 : Thu Oct 30 2008 - 17:45:20 CST