From: Antoine Leca (Antoine10646@leca-marti.org)
Date: Mon Mar 20 2006 - 07:43:57 CST
Philippe Verdy wrote:
> From: "Richard Wordingham" <richard.wordingham@ntlworld.com>
>>> I did not say that. I said that this support was not complete, and
>>> not working for all the announced Indic scripts that are
>>> configurable in IE options. If there's no way to configure it
>>> correctly for these scripts, these options should simply not be
>>> there.
Problem, some of us are able to make this support working (that is, using
proper fonts even if they are not always provided by Microsoft).
I am not saying it is always easy; it is often not, in fact. Just saying
that "no way" is far too extreme.
>>
>> I disagree. Imperfect support may be better than no support.
>
> Imperfect support that does not repect the language semantic and
> changes the expected visual order is worse than no support at all.
Hmmm, perhaps. Probably depends of the relative frequency of
"unrespectectful" support. The problem at hand is that I (or Peter) are
unable to reproduce the defects you are pointing out (for Gurmukhi), unless
I am going at very awkward configurations I know for sure are not working,
like using Office 2000 Arial MS Unicode, or on plain Windows 98.
> It gives users the feeling that this is the correct way to enter text
I disagree. You might have noted that the last thing Microsoft is releasing
to provide support for a given script, is the adequate keyboard layout.
If the base OS do not have the proper keyboard layout (such as the current
XPsp2 case of Oriya), you are not able to enter text, or if you succeed
(like using alternate input methods like BabelPad or Office's Insert a
symbol) you are not in the business of entering texts, just excerpts (this
is in turn useful for DTP tasks).
Designing a layout using KLM definitively requires skills and hence
knowledge of the script properties, so you then do not get wrong feelings.
Using precooked third-parties keyboard layout tools (like Tavultesoft's) do
not misguide either, since those tools come with documentation which
explains clearly what is working and what does not (I am thinking
particularly about Burmese here).
> that won't interoperate correctly with other systems, because users
> tweak their input to match the expected output, but break the
> standard logical encoding order.
I understand your point, but I fail to see how evident it could be.
If I consider Gurmukhi, which is one of the simpliest script, in addition to
the reordering of the i matra, it requires the handling of the -r and -h
subjoined consonants; I have yet to see a font or a rendering system which
provides the latter without the former.
If there is no way to create -r and -h, it is simply impossible to type
anything representative in Penjabi.
Another simple one is Tamil; yet, if the u and uu matras do not combine in
the proper glyphs, nobody will consider the resulting system suitable; and I
remain to see a system which combines u and uu but does not reorder the e/ee
vowels.
Antoine
This archive was generated by hypermail 2.1.5 : Mon Mar 20 2006 - 07:47:50 CST