* John Hudson | | Does anyone know how/if Opera 6 handles complex script processing?
What it does is basically to hand off the text to the OS, and let it do the processing and display. This works well, except for some issues with text selection, and, of course, bidi. | which could be interpreted to imply some limitations in Unicode | support. Any limitations found should be reported as bugs. There is full support for characters outside the BMP (except that you can select half-chars (reported as a bug)), matching of Unicode strings between HTML and CSS, upper-/lower-casing, legacy charset conversion etc etc. | What does Opera 6 do when it encounters an Arabic website? Nothing. It's displayed as if it the text were not LTR at all. On Windows NT and 2000 the OS will flip the text around inside each text node in the internal representation, creating a total mess. This is part of how Windows works, and I've been powerless to stop it. There is code for bidi support, but it's not entirely finished yet, and so was not enabled in the beta build. It will not be enabled in the final version, either. --Lars M.
