[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: TV tuner pci cards on FC2



Nigel Henry wrote:
> Thanks for that Peter. I think I'll go ahead and order that card. Did you see 
> that weird stuff after the "23.95" on my post. there was supposed to be a 
> Euro sign there which I got from my KDE's GB keymap using ALT-GR+4, like so.  
> €.

Worked fine here, and your e-mail correctly had UTF-8 character encoding.
Peter's response had the "€" [1], and a "Windows-1252" character
encoding.

UTF-8 encodes Euro signs as three bytes. I've just had a go on paper --
"€" is *precisely* what you get if you try interpreting a UTF-8
encoded Euro sign as three bytes in the Windows-1252 character encoding.

I'd guess that *something* in Peter's set-up doesn't know about UTF-8,
received a valid UTF-8 e-mail, decided it doesn't know about UTF-8, and
thought it ought to replace the "invalid" UTF-8 declaration with
something "valid". Something "valid" could be "US-ASCII", but none of
the Euro sign bytes are valid ASCII. It could be one of the ISO 8859
family, but the second byte of the Euro sign isn't valid ISO 8859. The
most common alternative is Windows-1252 -- given the number of
badly-written Windows programs around, it's certainly the most common
character set to be mislabelled[2]. So the program "helpfully"
relabelled the e-mail.

Obviously, this software has a massively exaggerated sense of its own
importance -- no-one would *ever* *dare* to use an e-mail standard it
didn't know about!

Once relabelled, the "€" interpretation is now "correct", according to
the modified e-mail.

Peter's using Thunderbird 1.5 (on Linux/Unix) as a client, and that
*does* know about UTF-8. So I'd imagine that it was something between
Peter and the rest of us.

Peter, do you use newsgroups to read the list? As far as I can tell,
newsgroup software and standards are lagging e-mail in accepting UTF-8.

If you're using standard e-mail to read and post, then there is
something in your system that is woefully broken. No e-mail software has
*any* business doing what I outlined above -- it is strictly against the
relevant standards, and will break on an increasing amount of e-mail.

Nigel, I'd go on using "€". This is an e-mail list for Fedora users, and
Fedora has always used UTF-8 as standard.

James.

[1] Note that this e-mail itself will be in UTF-8, so Peter won't read
those three characters correctly either...

[2] Microsoft "extended" ISO 8859-1 without bothering to give it a name
for years. So Windows programs thought they were sending ISO 8859-1
e-mail, and labelled it as such, when they were including non-ISO 8859-1
characters. The "‚" character is one of the extensions.

-- 
E-mail:     james@ | You can accept the existence of rain without denying the
aprilcottage.co.uk | existence of umbrellas.
                   |     -- http://ozyandmillie.org/2006/om20060615.html


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]