| |||||||||||||
|
| |||||||||||||
Issues with right-to-left languagesIssues with right-to-left languagesPosted Aug 29, 2002 17:26 UTC (Thu) by JoeBuck (subscriber, #2330)Parent article: GNOME Summary
While I'm at it: looking at the screenshot of the Arabic Gnome desktop, something strikes me as odd: Arabic, like Hebrew, is a right-to-left language. But in the Nautilus window, we see the Arabic text for "Location:" to the left of the URL. Now, I don't read Arabic or any other right-to-left language, but it strikes me that this is not natural; it would look to me as if I were expected to deal with a toolbar that says http://lwn.net :Location I would think that the user would want this to be flipped around, so that the label for the text field would be on the right.
(Log in to post comments)
It's worse than that... Posted Aug 29, 2002 17:45 UTC (Thu) by sab39 (guest, #2185) [Link] If your language goes from right to left, why would you think of an arrow pointing to the left is a useful metaphor for "back"?Now, I don't *know* that it would be wrong to use a left arrow in this way, but it seems like it would. I've heard that it's possible for either GTK or Glade to completely reverse the layout of your windows for RTL languages, and I'm surprised that wasn't activated here. For example, why should 'File' (the most commonly used menu item) be the last one to appear? Likewise, the Back and Forward buttons are the most frequently used in a browser, and appear last on the toolbar from this perspective. The tab controls in the dialog boxes appear at the end of the dialog instead of the start. Etc, etc. Perhaps the option *is* available and just wasn't turned on for the screenshot. I'd still be surprised (and impressed) if the sense of the 'back' and 'forward' button icons got reversed, though.
it's free software- finding bugs and fixing them takes time Posted Aug 29, 2002 19:12 UTC (Thu) by louie (subscriber, #3285) [Link] Gtk does have 'mirroring' ability, but... it's sort of a chicken and egg problem. The mirroring stuff (and other details that can't be taken care of at the toolkit level, but which must be done within the app) is not really tested much at all. Why is it not tested? well, there are virtually no translations, so it's hard for the people who would test it to use it. Of course, why are there few translations? Because there are many problems with mirroring and such. So... it's going to take some brave souls to test and find all of the bugs. Once that starts happening I expect we'll see fixes as well. [The Arabic team has already fixed several such bugs already, I believe, at least some of which will help all R-to-L languages.]
It's worse than that... Posted Aug 30, 2002 19:15 UTC (Fri) by Peter (guest, #1127) [Link] (Disclaimer: I do not speak or write any bidi languages.) If your language goes from right to left, why would you think of an arrow pointing to the left is a useful metaphor for "back"? Hmmmmmmmm, very good point. Your other points, about window tabs and so forth, are good as well. However, there is a practical issue: computers are predominantly Western, and UI development seems to be almost exclusively so. People who speak Semitic languages probably already use applications with window tabs on the left, <- arrows meaning "back", and so forth. While it may not seem natural to them in their native languages, it is a convention to which most are probably already accustomed. Changing the layout of an entire window just because it is being used in an Arabic locale would seem to cause continuity problems for those users migrating from less advanced software, or less advanced versions of the same software. Chicken and egg, indeed. If one can assume Semitic language users will predominantly be people new to computer GUIs, OTOH, your proposals make perfect sense. The mouse cursor itself should, of course, be mirrored as well. I'm gonna have to remember to ask my sister (a trained linguist who can read/write Hebrew) about this....
|
|
||||||||||||
