https://bugs.documentfoundation.org/show_bug.cgi?id=168537
--- Comment #14 from Eyal Rozenberg <[email protected]> --- (In reply to Khaled Hosny from comment #13) Khaled, I am myself somewhat partial to the view that direction should be considered an aspect of content rather than formatting. But making that choice is a very big deal - no less than making language an aspect of content (see bug 148257 and also some of bugs under meta-bug 162336). It means changes to ODF, it means taking direction out of page sequence styles, it means introducing meta-data entities at various levels, which is not the characters and breaks that we have now, and not the styles, but something different; it means separate UI outside of the dialogs and sidebar decks and sections which control Formatting; it means changes to import and export filters to accommodate this change; it means a migration path from ODF 1.4 and earlier and for individual documents; and I'm pretty sure there's some more work I'm forgetting. So, we can choose to do this. But given that we have not - we should not pretend that we have done it when considering one inconsistent command. I will also say that this is not a clear-cut choice, at all. First, think of languages like Japanese. In Japanese, you decide whether the text progresses LTR or RTL (or vertically). The glyphs can be laid out from left to right, or from right to left, and while the former is more popular - it is not uncommon to see the other choice (e.g. on both sides of a vehicle for symmetry). So, in that language it is really nothing but a formatting choice, not part of the content. As for languages like Arabic or Farsi - the characters themselves are RTL'ish, and that indeed is an aspect of content; but who is to say whether a given word or sentence in Arabic is within a paragraph that's RTL or LTR? The word itself will have the letters progress from right-to-left, but that's not actually something which we put into the styles. -- You are receiving this mail because: You are the assignee for the bug.
