I font to be readable

July 24, 2009 11:29 AM

Why do my Firefox and IE browsers, as well as iTunes, occasionally scramble a word or word part in normal text and instead display foreign characters?

Why do Firefox and IE browsers occasionally scramble a word or word part in normal text and instead display foreign characters, such as an A with a carat over it, or the Euro symbol, or the text string and% on some web pages? Also, when the Washington Post imports articles from other sites into their website, odd symbols appear in the article text, eg.: "Public health officials have said so-called ?social distancing? strategies ¿ sharply..." Finally, in iTunes podcast episode descriptions I frequently see the text string &md*ash (please mentally remove the asterisk, which I added in the middle of the string to force it to appear on this web page as it does in iTunes. Another example in iTunes: the word encyclopedie (with an acute accent diacritical mark placed over the middle e) is written as encyclo?pedie. What will fix this? Do I need to install another font? Thanks for your help.

posted by paphun123 to Computers & Internet (6 answers total)

Because it's TRYING to display the character, but your font does not have that character in it. This just means you need a better font, as they one you're using is crap

Try installing the DejaVu Fonts and using those.

The mdash thing is different. That is an HTML entity (a symbol that is used for web pages, to display special characters) - iTunes is not a web page, and doesn't know how to output those symbols the same way
posted by phrakture at 11:35 AM on July 24, 2009

Best answer: The A with a caret over it is a classic sign of character encoding mismatch: the page contains UTF-8 but the web server serves it with a ISO-8859-1 Content-Type header. A similar thing happens with smart quotes and the 1252 codepage. It means the person running the web site doesn't know what they're doing. Browsers tend to have a lot of heuristics built-in to try to guess the proper encoding of a page rather than trusting the incorrect hints given by the meta tags or http headers, but it doesn't always work.
posted by Rhomboid at 12:56 PM on July 24, 2009

The term for this is Mojibake or Krokozyabry. It basically boils down to non-matching text encoding schemes. If everyone just used UTF-8 things like this would happen far less often.
posted by zsazsa at 1:09 PM on July 24, 2009

Best answer: For more info, rather technical but readable for even a non techie, see Joel Spolsky's explanation. It's a hard problem to solve the right way, at least for programmers.

Don't lose heart though, that's just the problem facing programmers. You as a user should be able to get this to work right, at least most of the time. I'm sure other will post solutions, I just wanted to point out a good explanation that I read recently.
posted by intermod at 7:23 PM on July 24, 2009

Response by poster: Interesting article, intermod. Who knew plain text was so unplain!
posted by paphun123 at 12:23 PM on July 25, 2009

Response by poster: Just changed my default Firefox default encoding from Western ISO something-or-other to UTF-8. Perhaps that will help.
posted by paphun123 at 12:32 PM on July 25, 2009