On Sat, 17 Mar 2001, Marco Cimarosti wrote:
> But I see no easy way to conjugate the complexity of bidi embedding and
> overriding with the simplicity of a WYSIWYG representation. Not in plain
> text, however.
I do not see an easy way either. Good bidi editing, needs more thought and
experience. As I told before, you should have the resources to develop
different versions, and the courage to throw most of them to the garbage
can after you found that they're not good enough.
> If you allow embedding and overriding codes to be manually inserted by the
> user (as opposed as being automatically generated by the "DErenderer"), then
> you loose the simplicity of the approach.
I do not want my user even to know that there exist things named bidi
control characters. There should be mechanism for embeddings and
overrides, for example, a key that switches to right-to-left Latin
(da Vinci) or Greek. Also, an English word in the middle of Persian text
is surely an embedding, isn't it?
> But, although I mentioned rich text, what I really had in mind was plain
> text. Maybe in a rich text environment, where it is normal to get a run of
> text it and tag it with some property (e.g. underlined, bold, etc.), it
> would also be possible to select a block and say "this is embedded LTR
> text".
Perhaps the line between rich text and plain text is not that think
here...
> But, in this case, each *single* character in the block must be
> independently flagged with the property, so that it retains it also if it is
> copied&pasted somewhere else: the actual start and end codes will only be
> generated when rebuilding the Unicode string at the end of editing.
Now I like this. This is getting near to what I had in mind. Characters,
together with their embedding levels (and possibly more).
--roozbeh
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:20 EDT