Posted by Alan J. Flavell on 06/12/06 10:35

On Mon, 12 Jun 2006, Sid Ismail wrote:

> Have a look at http://www.elsid.co.za/symbols.html and tell me what I
> should do locally to be able to view the symbols as they _should_ be
> seen.

Initial impression is that it works. And that it has *nothing* to do
with your "Subject:" header!!!

Looking at the cited page
http://www.worldbridge.org/tourn/Verona.06/Bulletin/Bul0305.htm it
represents the suits by ♠ ♥ etc., which is technically
correct (might not be supported by some older browser/versions).

The page has no server-supplied charset specification, but it contains
the second-best substitute, a meta http-equiv, claiming iso-8859-1
(which, as we will see, is a lie, but doesn't affect the issue you
seem to be complaining about). Some older browser/versions (notably
NN4.*) would only be able to display this if the character coding was
stated to be utf-8, but that's a browser bug: the page in question is
OK.

HTML validation shows that it contains some defective characters, but
they aren't the suits: see
http://validator.w3.org/check?verbose=1&uri=http%3A%2F%2Fwww.worldbridge.org%2Ftourn%2FVerona.06%2FBulletin%2FBul0305.htm
which contains some advice on how to correct those errors. Changing
the character encoding to windows-1252 would be a quick workaround,
but it would be better not to rely on a proprietary encoding.

Coming back to http://www.elsid.co.za/symbols.html , it has neither
HTTP-specified nor meta...http-equiv specification of character
encoding, so all bets are off.

The validator guesses that it's utf-8 and refuses to even try to
validate it, which is technically correct. If I force iso-8859-1 and
HTML/4.01 transitional, there are still far too many irrelevant errors
to be able to spot the real ones, but you still have at least one
Windows-specific (non-iso-8859-1) character in there.

> My thanks in advance!

Well, there's quite a bit wrong with both pages, but none of the
faults seem to be quite the one that you're complaining about, and the
"Subject" header is still **VERY** misleading, as there is no use of a
"Symbol encoding" nor a Symbol font in here (which would anyway be
bogus in proper HTML).

[Back to original message]