Thanks for your quick reply guenter. Sorry to bug again, but setting to UTF16 doesn't work for me, in fact it breaks the page displayed completely... just array of gibberish without any formatting. I understand your 'a gecko is a gecko' argument too... that's why I wish to fix this problem if possible.
Using this link again (I hope you don't get tired of it)
http://udn.com/NEWS/main.html
(I am quite certain this site uses Big5 encoding exclusively)
I have tried the following:
Firefox 3.0 English stock
Autodetect -> Big5
Force Big5 -> OK
Force UTF16 -> breaks page completely
Force ISO-8859-1 -> gibberish
(Basically, no surprises)
Now, K-Meleon 1.1.5 English stock
Autodetect -> Big5, text OK, but layers and title is ????
Force Big5 -> same as autodetect
Force UTF16 -> breaks page completely
Force ISO-8859-1 -> text gibberish, BUT layers and title OK!!
A surprise. You might remember that I had set Windows non-Unicode programs to display as locale Chinese(Taiwan), which means Big5. So now I set the Windows non-Unicode locale back to English(US).
And try K-Meleon again:
Force Big5 -> text OK, layers and title is ???? (same as before)
Force ISO-8859-1 -> text gibberish, AND layers and title gibberish!!
In summary, to display correctly:
body text -> K-Meleon force to Big5, or autodetect
layers and title -> K-Meleon force to ISO-8859-1, let Windows handle it as if Big5?!
Are the layers/title and the body text following identical codepage<->unicode conversion paths? If they are, did I make a blunder somewhere?