The Unicode Consortium Discussion Forum

Hi all,

I would like to propose an algorithm in order to permit display of language independent Boustrophedon text (http://en.wikipedia.org/wiki/Boustrophedon). Boustrophedon text reverts (mirror vertically) characters and change direction on each line.

The proposal would be the following:
> The unicode characters:
- LRM (U+200E) LEFT-TO-RIGHT MARK and
- RLM(U+200F) RIGHT-TO-LEFT MARK
> when applied on strong directional characters (e.g. R),
> they force the glyphs of the character to be mirrored vertically
> and they set a new direction of text.
> Example [R+RLM+A+RLM+!+RLM] will display [!AЯ]
> The algorithm should not require new encoding for mirrored glyphs but requires computer software to handle this mirroring automatically.
> Possible extension/change would be to have a unicode character that marks all the following text as Boustrophedon and let the computer software handle carriage return and text layout .

May I have your thoughts about that.

Many thanks in advance,
Samuel

I know that a Boudi specification has been considered before, but it was decided that it more properly belongs to the domain of rich text. The only thing that has really changed is PRI 207 for UTR #50, which defines properties for vertical text layout. If you could get an agreement to add a boustrophedon text style to HTML/CSS from W3, or the Open Document specification from OpenOffice, then I think Unicode would have cause to define properties for boustrophedon text.

I definitely would vigorously fight against overloading the RLM/LRM and other BiDi controls, as it would seriously mess with the Bidi algorithm. If you need to control mirroring for individual characters and runs of characters, then I would suggest separate Inhibit Mirroring, Force Mirroring, and Mirror Override control characters, but a proposal for those sorts of things will only be considered by the UTC if we know that they are actually necessary.