I posted this as a comment but I want to start a new thread because I think it's a unique problem.
I'm using WordPress 1.2.1 for a website about the war situation in Nepal. I can type Indic/Devanagari script posts in the WordPress post text area, or I can cut and paste from another source. Anyway, I am posting valid Unicode Devanagari script of Nepali texts.
It shows up properly in MS Internet Explorer, but not in Mozilla. But it does show up fine in the WordPress post page, and on the preview on the lower part of that page.
In other words, it's stored fine, and Mozilla is able to show Indic Unicode fine. But WordPress must be doing some mangling processing when it shows the post on the main page.
You can see a sample here:
(If you don't know what good Devanagari looks like, it doesn't have all those dotted-line circles. Those are some kind of placeholder for some Devanagari characters, and they should not show up in the final text.)
As an old-time coder, my suspicion is that some algorithm is taking the post and spitting it out character-by-character with some invisible spacer in between, and this is breaking up all the character combinations. In Devanagari, characters show differently depending on what is immediately before or after it, because they combine together. If they are broken up then they show up as free-standing versions with that dotted line circle as placeholder.
I even tried a plug-in called wpunformatted that I found on the net, but it didn't change anything.
How can I stop WordPress from mangling my beautiful Nepali texts?