Re: [Ietf-languages] Fwd: I-D Action: draft-msporny-d-langtag-ext-00.txt

[Richard Wordingham <[e-mail]>]

On Mon, 27 May 2019 18:24:56 -0600
"Doug Ewell" <[e-mail]> wrote:

> Richard Wordingham wrote:
> 
> > So how do you use a present-day language tag to set the paragraph
> > direction for Middle Egyptian in hieroglyphs?  Do you have to do
> > something like ensuring that the scope of the Middle Egyptian
> > language tag doesn't extend to the starts of paragraphs?
> > (Reminder: Ancient Egyptian Egyptian hieroglyphs could be laid out
> > in either order; right-to-left was the default, but nowadays the
> > default is left to right, apparently even in Arabic publications!  

> My arguments regarding the proposed D extension have been that:
> 
> (a) for the posited use cases, primarily Arabic, the proposed
> solution is not necessary, because explicit or implicit script
> subtags already indicate direction, and
> 
> (b) for use cases such as mine, the proposed solution is not
> sufficient, because the repertoire is not extensible.
> 
> > And what if you want to have English text in right-to-left
> > paragraphs?  

> I assumed we were sticking to real-life use cases. That said, in
> BabelPad I get the desired result with a Latin-script line by
> prefixing it with RLO.

Surely RLO would result in English being written backwards?

‮As here!

I believe that may result in right-to-left direction (as with my
email cient) because there is no overriding protocol, so first strong
character determines. In normal text, one can achieve that effect with
RLM or ALM if there is no contradictory higher level protocol. The most
striking effect that I can think of on purely English text is that
'unjustified' text typically ends up flush right rather than flush
left.

There was a period when LibreOffice used to misinterpret my
keystrokes as commanding a change of paragraph direction to
right-to-left.  The result was misery as I searched for the command to
undo the uncommanded change of direction.

Richard.