Comments on: Character encoding basics

By: Emil Stenström

Emil Stenström — Wed, 22 Nov 2006 22:32:23 +0000

@Rasmus Kaj: Thanks for the clarifications :)

By: Rasmus Kaj

Rasmus Kaj — Wed, 22 Nov 2006 22:12:51 +0000

Actually, at first computer people thought 128 characters to be enough, and thats what ASCII is. The eight bit was used for checksumming or (often) not at all. There was localized versions of the 7-bit character set, such as the one replacing [|]{\} with the Swedish åäöÅÄÖ.

After a while, to important steps were taken; 1) The eight bit was used, doubling the number of characters, and 2) ways of specifying the character set in (mainly) email started to develop.

From that, the story goes pretty much as you tell it … :-)

By: Nita

Nita — Fri, 27 Oct 2006 17:10:33 +0000

“So, rely on your server to send the right encoding.”

That’s easy to say, but not everyone has the power to set server headers ;)

Here’s a fun article on the inner workings of character sets, if anyone’s interested: http://www.joelonsoftware.com/articles/Unicode.html

By: Emil Stenström

Emil Stenström — Fri, 27 Oct 2006 16:10:29 +0000

@Mark McDonnell: That means that you don’t send *any* encoding at all (not that uncommon). Have a look at the example code for you language and add it to your main template. No need to keep the browser guessing what encoding you use.

By: Mark McDonnell

Mark McDonnell — Fri, 27 Oct 2006 12:47:41 +0000

Strange, I used the “View Response Headers” on one of our websites:

http://www.firetuck.co.uk/Home.asp

…and found that the response headers didn’t show anything for Content-Type OR Content-Encoding.

Content-Type did show “text/html” but nothing about the character encoding.

Any suggestions?

By: Birgit

Birgit — Sat, 21 Oct 2006 09:36:55 +0000

Hey, thanks a lot, this really helps! I never knew about the “View response headers” option and never understood why I have to encode certain characters although I set the encoding in the HTML header…now I know :)

By: Emil Stenström

Emil Stenström — Sat, 14 Oct 2006 08:53:07 +0000

@Juani: Yeah, I have some experience with manually maintaining local languages with entities.UTF-8 is just way better :)

By: Juani

Juani — Thu, 12 Oct 2006 02:17:27 +0000

Hi Emil! Thank you very much for this article! Until now I was manually/automatic encoding my chars (I develop in spanish) and was really painfull, now I only have to save it in UTF-8… simply brilliant!

By: zcorpan

zcorpan — Wed, 11 Oct 2006 22:49:28 +0000

Python is without the semi-colon. ;-)