On Sun, May 19, 2002, Tzafrir Cohen wrote about "Re: official hebrew in Linux-IL mailing lists?": > > > They use different *implementations* of the same algorithm > > > (KDE: QT's implementation, Gnome and abiword: fribidi, mozilla and > > > openoffice: ICU) > > > > the point is that they would show some sentances not in the same way.. > > No. The unicode specification specify exactly how a compliant > implamantation should display bidi (convert logica->visual). All > implementations should be strictly compliant. If two separate > implementations display the same sentence in a different way, then at > least one of them is buggy.
Almost, but not exactly. The problem is that Unicode doesn't specify the major direction of the text. A plain-text file doesn't have any direction queues besides the text itself (assuming it doesn't contain special LRM codes, etc.) so when I wrote bidiv (for example) I found myself having to invent heuristics (paragraphs seperated by an empty lines, etc. - we talked about that in ivrix-discuss). What if my huristics are different from those used by Microsoft? What heuristics do the KDE or Gnome mail programs use when showing a Hebrew/English email to determine the main direction of a line? I guess that to be really sure how your text is displayed, you must start each sentence with a unicode direction character. But these are not available in ISO-8859-8-i, which most Hebrew email and text files have been written with so far. -- Nadav Har'El | Sunday, May 19 2002, 9 Sivan 5762 [EMAIL PROTECTED] |----------------------------------------- Phone: +972-53-245868, ICQ 13349191 |We aim to please, you aim too, please. http://nadav.harel.org.il |(sign in a gas station men's room) ================================================================= To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
