Re: [emacs-bidi] status? news?

[Roozbeh Pournader]

On Sun, 21 Jan 2001, Eli Zaretskii wrote:

> > Perhaps they're doing this to stop us from developing our good programs :)
> 
> That's not funny: it might be one of the reasons.

I was thinking about that. But that's really a sad thing to think of. So I
made a joke.

> I don't think your fears are justified.  The bidi algorithm (_any_ bidi 
> algorithm) is table-driven.  All I was talking about was slight changes 
> in the tables which drive the algorithm.  Problems that result from
> incorrect tables are therefore easy to fix: just change the tables.
>
> Such problems are also easy to spot: since any given character 
> category includes more than one character, you don't need to wait for 
> long before the display shows incorrect results: they really stand out 
> for someone who knows the language.
> 
> Something that is easy to spot and easy to fix should not scare us, I 
> think.

You were talking about introducing new types. And you're talking about
using common sense to fix the bugs. Because of the difference in common
sense of bidi users, even when they use the same language, they expect
difference things.

If we're doing an extension to the Unicode behaviour, we should try to
standardize it in an RFC or something, and we should start two separately
developed versions to test against each other.

--roozbeh