Hi, I am using debian and I wanted to install some unicode fonts. When I did a "apt-cache search" I found quite some packages which seem to be unicode fonts. However, most of them will tell you that it is a unicode font for chinese, greek, etc.

This make me wonder about this question on how softwares handle unicode fonts. Let's say if a webpage/text file was written in Chinese and Greek. Then do you need a unicode font which contains both of these 2 languages to display it? Or the system is clever enough to find the characters from different font files?

Thanks for your enlightenment in advance

I havn't looked into this, so it's speculation...

<speculation-mode>
Unicode is LARGE and needs lot of fonts to present all of its characters. So to save room on your disk, they have splited it into parts. So you download the parts you will use.

In Emacs, characters that isn't possible to show is shown with a square. I think it's the same with other programs, like FireFox. I have hade those problems here with Firefox (see my name, and you will understand and problem with page coding from server/pages, not Firefox)
</speculation-mode>

Thanks for your reply.

But if that is the case, does it only use fonts specially made for unicode? Or does it can use any usual TTF files?

How fonts are visually shown has nothing to do with character coding.

So a character 'A' coded in Latin1 (ISO-8859-1) has nothing to do with which font it is used to shown the 'A* on your screen (or print out).

So an 'A' in Unicode and an 'A' in Latin1 can look the same or not, depending on fonts (looks). But it's still the same character, hopefully. There are at least two 'Å' in Unicode. One for Swedish/Norwegian character 'Å', and one for the length ångström (used with atoms, so it is a short unit), which is named after a Swedish physics scientist. I guess they didn't ask anyone from Sweden or Norway about that

Anyway, you have to remember that what you see on the screen is a font, that represents graphically a character in a character coding (like Latin1 or Unicode or something else). So you could use many different fonts to represent the same character.

So if those fonts you download to be used in Unicode character coding could be in any font format that your system can handle, like TTF (True type font). So I think that there is some sort of mappings between Unicode and other character codings that different fonts are coded in.

Confused? I'm not that convinced by my own explanation

Indeed
But I more or less understand what you meant.

Thanks for your lengthful explanations though

Thanks.
Your welcome. Just hope someone else get any use of it.