Ignacio Renuncio <[EMAIL PROTECTED]> wrote: > The XML has no charset errors, it's encoded in UTF-8 and is > charset-validated. But it contains such special characters that cannot be > converted to ISO-8859-1. It's the same thing that happens if you try to put > Chinese text.
That's what I said. So, I would recommend not converting it to ISO-8859-1, since it is not possible. > I think the problem could be related to: > > A) The browser, the autodetected charset used to display the document could > be wrong. Most modern browsers have no problems with this. You could try explicitely switching it to UTF-8, which sometimes helps. > B) The font: Could be unable to display some characters. That could be it. In such cases you would normally see one square, or one question mark or so. If you see several characters of garbage, than there's ussually somewhere on the track from the xml to the browser a misinterpretation UTF8 as iso-1. > C) My_editors uses ISO-8859-1 so the conversion from UTF-8 is not possible. In this case you should see question marks for the unknown characters, because the conversion is done by java, and it produces those signs for impossible conversions. > > Should I take it like a "feature" or a "bug"? That depends one what is the problem. > Chinese users, do you hear me? How did you cope with this? No need for chinese users. An average quote from Word is already enough to go out of the scope of iso-1. See also the the 'encoding' example, which contains several examples of non-western-european characters. Are those displayed correctly in your browsers/editor? > I'm only using UTF-8, the XML file is encoded as UTF-8 and its characters > are 100% legal. Maybe the "my_editors" output charset is set to ISO-8859-1 > and thus cannot cope with an impossible conversion. How about basic jsp editors? I'm sure that they are UTF-8 (and btw, even have a more-or-less spanish interface available :-) Michiel -- Michiel Meeuwissen Mediacentrum 140 H'sum +31 (0)35 6772979 nl_NL eo_XX en_US mihxil' [] ()
