Build: 05-01 branch build. Steps: 1. Go to a page that has links to non-ascii characters in NCR - I'll post a test case later. 2. Move the mouse cursor over the non-ascii links. 3. Look at the status bar. Result: Those characters are not in the page encoding are displayed garbled. IE show them all correctly (in Unicode). This might related with bug 92109, but that one is for when we input non-ascii character in URL location bar, and they are displayed as escaped characters.
Created attachment 81951 [details] a test case This is a page in gb2312, same problem can be reproduce on other encoding like shift-jis. Move the mouse on the top of links one by one to see the difference.
I am not sure if we could display unicode string in status bar directly. If we are using system API, that will be a problem. But at least there is another proble. nsStandardURL is implemented using native charset, and that does not work with characters that are not in native charset. Some big change needed to make it work. (Run the test case. Check output console, and you will see message like: WARNING: unicode conversion failed, file y:\mozilla\netwerk\base\src\nsStandardU RL.cpp, line 111 Set a breakpoint there. )
Yuying, can you try to reproduce this with a current trunk build? It works for me with 2003010705 on Win2k.
on 01-06 trunk build / WinXP, all the letters are displayed properly in status bar except the 6th one "水" is showed as "?".
All the testcases now work.