From: Omer Zak (omerz@actcom.co.il)
Date: Sun Sep 22 2002 - 19:08:19 EDT
On Sun, 22 Sep 2002, Andrew Dunbar wrote:
> I think the first and most important step here is to
> build avery full and descriptive list of all the
> quirks, and hopefully decide what a 100% quirk-free
> bidi algorithm would be.Is the Unicode bidi
> algorithm
> perfect if everybody implemented it perfectly?
It is not perfect (at least according to Arab/Persian speakers), but it is
the Standard.
And there are Unicode 3.0, Unicode 3.1 and if I am not mistaken, also
Unicode 3.2. The Law is the latest Unicode, and the previous standards
should be supported as quirks.
In other words, we need to provide in AbiWord document format for a field,
which will specify the BiDi algorithm under which it is to be rendered.
This is to allow old documents to be correctly rendered, even if the
standards, upon which they are based, have been modified.
> Is everybody going to agree on what it should be?
No problem, Unicode is the Law.
> ... I
> hope so.It sounds like each quirk should be assigned
> a flag which freebidi can query and import/export code
> should set these flags according to the format
> (hopefully we'll not have to take in user prefs for
> things like this if there is truly only one right
> way).
I agree with this approach. When saving a document, it may be a good idea
to give the user the option of either continuing to save it in the quirky
format, or to reformat it so that it'll be rendered the same under the
current standard. This may not be possible for all quirks, but for some
of them, it should be possible.
> I think I've seen these kinds of quirks in Hebrew with
> hyphens connecting Hebrew endings to English words.
> I got varied results between different Unicode
> editors,
> mozilla, and IE; and more variation depending on
> whether I used theASCII hyphen or the special Hebrew
> hyphen.I think World.abw contains this example.
>
> Is this one of the known quirks?
Yes.
--- Omer
WARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html
This archive was generated by hypermail 2.1.4 : Sun Sep 22 2002 - 19:12:53 EDT