By using this site, you agree to our updated Privacy Policy and our Terms of Use. Manage your Cookies Settings.
440,262 Members | 1,125 Online
Bytes IT Community
+ Ask a Question
Need help? Post your question and get tips & solutions from a community of 440,262 IT Pros & Developers. It's quick & easy.

Slightly off-topic question

P: n/a
Ok sorry for the off-topic question.

I see a lot of people posting their html code here.
Why so many people still use iso-8859-X instead of UTF8? Is there a
particular reason?

For my latest develops (web pages in greek) I always use utf8 both in
html&DB. Isn't that the right thing to do?

again sorry for the off-topic question
Dec 22 '06 #1
Share this Question
Share on Google+
11 Replies


P: n/a
Harris Kosmidhs wrote:
I see a lot of people posting their html code here.
Why so many people still use iso-8859-X instead of UTF8? Is there
a particular reason?
Laziness, in my case. Dreamweaver uses 8859-1 by default. I only
changed it recently, when I found out more about 8859 and UTF, and also
found out that it was relatively easy to change Dreamweaver's default
too.

--
K A Nuttall
www.yammer.co.uk
Re-type the e-mail address how it sounds, remove .invalid
Dec 22 '06 #2

P: n/a
Scripsit Harris Kosmidhs:
Ok sorry for the off-topic question.
You are not excused. Learn how to find the right group and how to write
meaningful Subject lines. "Off-topic" doesn't belong there; either your
message is on-topic, or it should not be sent.
I see a lot of people posting their html code here.
That's a mistake, as the regulars regularly explain, except perhaps in
hopeless cases.
Why so many people still use iso-8859-X instead of UTF8? Is there a
particular reason?
It doesn't matter from the CSS perspective, since CSS should normally use
only ASCII characters anyway, to be on the safe side. CSS is normally not
viewed by users who visit web pages, so there's no compelling reason to
localize it.
again sorry for the off-topic question
Stop babbling that. It doesn't make the off-topic posting any better.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

Dec 22 '06 #3

P: n/a
On 2006-12-22, Jukka K. Korpela <jk******@cs.tut.fiwrote:
[snip]
>Why so many people still use iso-8859-X instead of UTF8? Is there a
particular reason?

It doesn't matter from the CSS perspective, since CSS should normally use
only ASCII characters anyway, to be on the safe side. CSS is normally not
viewed by users who visit web pages, so there's no compelling reason to
localize it.
What about values of the content property (used with the :before and
:after pseudos)?

You might need those in Arabic, Chinese, etc.

I'm guessing that content language and encoding should be delivered for
CSS files in the server headers in just the same way as for HTML and
script files.
Dec 22 '06 #4

P: n/a
Scripsit Ben C:
What about values of the content property (used with the :before and
>after pseudos)?
Their usefulness in WWW authoring is very limited, due to lack of any
widespread support.
You might need those in Arabic, Chinese, etc.
Might I? I doubt that. But even then, the strings would typically be short,
e.g. specifying quotation marks or markers or whatever, and using the \u...
escapes I could still have the CSS code written in ASCII only.
I'm guessing that content language and encoding should be delivered
for CSS files in the server headers in just the same way as for HTML
and script files.
Content language is immaterial. Encoding is specified, if needed, in HTTP
headers as for any other subtype of text. The rest is a sad story of
confused guesswork, but we don't need to let things go there, do we?

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

Dec 22 '06 #5

P: n/a
On 2006-12-22, Jukka K. Korpela <jk******@cs.tut.fiwrote:
Scripsit Ben C:
[snip]
>I'm guessing that content language and encoding should be delivered
for CSS files in the server headers in just the same way as for HTML
and script files.

Content language is immaterial.
For CSS files, yes; for HTML documents it influences font selection, and
it would be nice if more webservers delivered the Content-language
header so browsers didn't have to guess the language from the encoding
or from the actual characters present (or various other tricks).
Dec 23 '06 #6

P: n/a
Ben C wrote:
For CSS files, yes; for HTML documents it influences font selection, and
it would be nice if more webservers delivered the Content-language
header so browsers didn't have to guess the language from the encoding
or from the actual characters present (or various other tricks).
The content-language header just describes the primary language of the
intended audience. The HTML lang attribute describes the language the
document is actually written in.

--
David Dorward <http://blog.dorward.me.uk/ <http://dorward.me.uk/>
Home is where the ~/.bashrc is
Dec 23 '06 #7

P: n/a
Scripsit Ben C:
>Content language is immaterial.

For CSS files, yes;
Which is what we discuss here.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/
Dec 23 '06 #8

P: n/a
On 2006-12-23, David Dorward <do*****@yahoo.comwrote:
Ben C wrote:
>For CSS files, yes; for HTML documents it influences font selection, and
it would be nice if more webservers delivered the Content-language
header so browsers didn't have to guess the language from the encoding
or from the actual characters present (or various other tricks).

The content-language header just describes the primary language of the
intended audience. The HTML lang attribute describes the language the
document is actually written in.
That too, yes, but if there's no lang attribute, the Content-language
header is also a good place to look.
Dec 23 '06 #9

P: n/a
Ben C wrote:
>The content-language header just describes the primary language of the
intended audience. The HTML lang attribute describes the language the
document is actually written in.
That too, yes, but if there's no lang attribute, the Content-language
header is also a good place to look.
Not that good, just because a document is written for people who's primary
language is English doesn't mean that the document is written in English.

.... and if an author has failed to use the lang attribute, how could the
Content-Language header be trustworthy? Its much harder to set up.

--
David Dorward <http://blog.dorward.me.uk/ <http://dorward.me.uk/>
Home is where the ~/.bashrc is
Dec 24 '06 #10

P: n/a
On 2006-12-24, David Dorward <do*****@yahoo.comwrote:
Ben C wrote:
>>The content-language header just describes the primary language of the
intended audience. The HTML lang attribute describes the language the
document is actually written in.
>That too, yes, but if there's no lang attribute, the Content-language
header is also a good place to look.

Not that good, just because a document is written for people who's primary
language is English doesn't mean that the document is written in English.
Not a bad point, but W3C do say you can use it:

http://www.w3.org/International/questions/qa-lang-why

"User-agents can (and do) use the content language to select
language-appropriate fonts, which improves the overall user
experience of the page."

And in the absence of a lang attribute, it's worth a try.
... and if an author has failed to use the lang attribute, how could the
Content-Language header be trustworthy? Its much harder to set up.
Not trustworthy, but better than nothing I suppose.

There's a similar situation with encoding: there it's actually specified
that if there's a metatag in the content and a server header saying
different things, the server header wins.

I'm not sure what the reason is for that priority order-- perhaps it
_is_ assumed that the server configuration is more likely to be set up
correctly (even it might be harder) than that the content is correct in
these details.

After all the server for a big site is set up infrequently, but new
pages appear every day, written by web authors and we know what they're
like.

The other reason may be that since for many encodings you can't even
read the metatag unless you know the encoding first, the metatag is not
a sensible place to put the encoding.
Dec 24 '06 #11

P: n/a

Harris Kosmidhs wrote:
Why so many people still use iso-8859-X instead of UTF8? Is there a
particular reason?
IMHO, ignorance, followed by legacy issues.

I author UTF-8 pages at present that contain characters I'd otherwise
need to use three different and incompatible ISO-8859-* encodings to
represent. I can't practically work without UTF. These pages are Java
generated and one set of JSP code supports all countries and character
encodings. Again I need to Unicode to do this without a lot of
additional customisation on each page.

There's still an attitude with some developers that Unicode is either
"hard" or "inefficient". This leads them to think that non-ASCII
solutions require ISO-8859-* encodings, and of course these mean
different encodings for different regions, even though they can
potentially work "anywhere" (one at a time anyway).

I've also got one legacy-ish customer where their Arabic Oracle
database runs under Solaris, yet has been configured for a Windows
codepage (Arabic)! Amazingly this manages to hang together throughout
the whole system, but life would probably be simpler overall if we'd
stuck to UTF-8 throughout.

Dec 27 '06 #12

This discussion thread is closed

Replies have been disabled for this discussion.