> The purpose of UTF-8 was to allow everyone to use their language characters in filenames (among other things, of course). Moodle already supports UTF-8, which is a great thing.
If your people use UTF-8 characters in file names and you all don't have problems, why not?
Let me describe my environment. Switzerland is highly multilingual. They have four official languages, in my area German is the first language, and French is the second. So we get all ä, ö, ü and ß from German and a whole bunch of characters like à, é, è, ô. And peoples names can have anything from ISO-8859-1 or -15.
I think because they are multilingual they are used to work with different alphabets and have much less problems in understanding that computer has its own alphabet. André Müller knows that his e-mail is email@example.com or, in the worst case, firstname.lastname@example.org! And I don't think they try to save a digital recording of a song in http://en.wikipedia.org/wiki/Romansh_language
in its Rumantch name!
On the other hand, files with special characters in their names might "work" in one computer system but not in another. Just an hour after my previous post, I tried to backup a huge directory from my ext4 file system to an external USB drive on NTFS and got endless "wide charactor" errors. The reason, collegues using other OSes have used ä, ö, ü in file names. Somehow I could transfer them to my Linux system, I guess either through Samba or HTTP. Now before moving them to the external drive, I have to rename them. The greatest adventure I had was with a file name starting with "spaces" - yes, there were two of them!
> So, the effort taken to give computers the capacity of "knowing" how to deal with almost all characters, in different types, is indeed a good thing.
That is exactly what I feared. It is not UTF-8 but a subset of it. Is it clear to everybody what that subset is?
Look at this: http://moodle.org/mod/forum/discuss.php?d=171081
. Guillermo M. reverse engineered the 7-bit ASCII set to find out the subset of special characters allowed in passwords. Imagine repeating the exercise for UTF-8!
> But it should not be necessary for us to adapt to computers: they have the computational power to be adapted according to our needs - and that's how it should be!
Considerate computers? I'll be glad to meet them.
P.S. Since version 2.0 Moodle has a new way of saving files. Unlike 1.x the file system of the server
does not see the "real" name of the file. The real name is just a database entry. So Moodle make you to believe a fancy file name (this may not be accurate, I still don't have 2.x in a production system). So main point of this discussion is general, "should one use the whole UTF-8 character set in file names".