For producing: Go to Google, write "tIrtIl Seker" (Not, you may need a turkish keyboard letters "I" in "tIrtIl" is actually "i" without point, letter "S" in "Seker" is actually a "s" with a point down, read as "sh") in the search form area. When the result are come you see the search words as "tırtıl şeker" in search form area. i believe it is releated with Bug 117422 Char set: ISO8859-1 Mozilla/5.0 (Windows; U; Win98; en-US; rv:0.9.9+) Gecko/20020319
Ahmet, If you try the same thing on IE (Encoding : Western ISO) you get the same result, I guess the reason is, if you submit Turkish character on a web site with default encoding is non-Turkish you get converted characters. So, first change your encoding to Turkish then submit the form, you will get the correct result. But I dont know whether this behaviour is correct. Should this bug be marked as "invalid" ??
And another point is , if you do such a search in google, it does the correct search, you just lost the characters if encoding is non-turkish.
Very similar with bug 125034, but that one is linux only. cc Shanjian.
there are no such turkish characters in ISO-8859-1, change it to ISO-8859-9 and you should be able to work mark this as invalid. this is a user error
Marking as verified per comment above.