when i browse with firefox it happens that - sometimes, not always - special signs from the norwegian or german alphabet are displayed incorrectly as a question mark, like this: http://home.arcor.de/bergenbilder/signs.jpg
i use the newest version of firefox and think that i'm running the newest unicode, too - is it a problem at the other end!? any possibilities to change this?

regards
ungua

Can you give me a link to that site / dialogue box, so I can test it on my system?

Robin

http://mail.yahoo.no - screenshot
http://web.de - german webportal

Ok, I've taken a look at the source code of the Yahoo page. The question mark appears in two places where an a with a circle above should appear, which seems to be typed directly into the source code and not with the appropriate HTML code; all the other question marks appear in places where the code shows the HTML code for 'non-breaking space'. I don't know why that is, but you should try another version of your browser, or a different browser altogether. The problems with non-breaking spaces hint that there's nothing wrong with your fonts, it seems your browser cannot interpret the HTML-coded symbols.

Robin

okay, thank you very much! i need to use mozilla-firefox because konqueror won't be able to open my gmail-account. additionally, konqueror freezes quite often. but it's good to know the problem doesn't appear to be serious...

regards
ungua

p.s.: in windows i could use the ascii-code with typing in numbers that were compiled into characters. now i have the norwegian signs like "Ã¥" but not the swedish/german signs. how do i solve this? is there any faster way than shifting keyboard-layouts?

I'm not sure how to enter those symbols, actually. But maybe the following article can help you:

http://www.linux-user.de/ausgabe/200...78-answergirl/

Scroll down until you get below 'Kasten 1', where there is a description of some graphical tool for the editing of keyboard layouts. The Scandinavian a-with-circle-above is actually used as an example there.

Robin

that sounds interesting. i've scanned the text and have a try later - thank you!

regards
ungua

after updating to mozilla 0.9 i never had this problem again. but today i read two different texts from the same source, one with correct display of signs (a), one with wrong display (b).

a: http://www.ssb.no/samfunnsspeilet/utg/200204/10/
b: http://www.ssb.no/samfunnsspeilet/utg/200005/4.shtml

i can have them open at the same time and see one with and one without mistakes. how can this be explained and even more important, fixed!?

regards
ungua

Hi,

I also have sometimes this problem.
It seems to me to stem from the encoding scheme used by webserver, webdesigner and webbrowser.

Please take a look at the HTML-Sourcecode.

There shoud be no difficulties if the special characters are entered as HTML-Entities (ü as german "umlaut" ü).

The second possibility is to state the encoding you use in the HTML-Source and then use the normal Characters from your machine in HTML-Code.
<meta http-equiv="Content-type" content="text/html; charset=UTF-8" />
Now the Browser should use right encoding scheme.

The third possiblity is to rely on the webserver transmitting the right encoding and not care about the HTML-Document at all. This seems to be the reason for most improper websites, as SuSE-Linux uses UTF-8 encoding, while "Windows" and often even "Unix" use different.

Please correct me if I am wrong about this subject.


Puzzled Pollyanna

hi...
i didn't really understand your options - du kannst mir aber auch auf deutsch schreiben, vielleicht fællt mir das leichter... schreibe gerade an meinem examen und die beiden seiten stammen ja von einem server, das verwundert mich. und bei yahoo funktionieren die zeichen manchmal, manchmal auch nicht - mir vollkommen unverstændlich!

lieb gruss
ungu

Hi,

is my English that bad? ;-)
It seems to be like I mentioned it before:

In a)
http://www.ssb.no/samfunnsspeilet/utg/200204/10/
which shows propper Norweignian Characters in the beginning of the HTML-Code the encoding schem is propperly defined as
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
which is to my knowledge typical for Linux/Unix.

In b)
http://www.ssb.no/samfunnsspeilet/utg/200005/4.shtml
there is *NO* encoding scheme in the HTML-Code, so we all depend on a usefull webserverconfiguration, which does not seem to help very much, if we both are using SuSE 9.1 or 9.2, which relyes on UTF-8 as encoding scheme.
If ther is no or wrong character-encoding submitted by the webserver, you do only see this error if you use a different character-encoding-schem on *your* machine than the author on *his* machine. If you use SuSE 9.1 or 9.2 this is very probable, because since SuSE 9.1 the default encoding scheme is UTF-8, wich I think of as a good choice.
If the webauthor has written his text encoded in ISO-8859-1 (Linux) or even Windows, and did not bother to use HTML-Entities for his special characters as it is wildly used today, *we* do see only rubbish of the special characters if we try to look at this document in UTF-8 encoding.
Change your character encoding scheme of Konquerer from "automatic" to "manuelly: ISO-8859-1 an all special caracters on site b) are displayed correctly.

I hope this was understandable.

Good Night
Pollyanna

and now in German.


Hi,

ist mein Englisch wirklich so grauenhaft? ;-)

Es scheint so zu sein, wie ich es vermutet hatte:
Auf der Seite a), die *korrekte* norwegische Sonderzeichen anzeigt, ist im HTML-Code der Zeichensatz mit
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
ordnungsgemäß definiert. "ISO-8859-1" ist übrigens eine Linux/Unix-Standard-Zeichenkodierung.
Hier hat der Autor der Webseite einfach die Sonderzeichen ganz normal heruntertippen können, wenn er auf einer Linux/Unix-Maschine mit ISO-8859-1 als Zeichenkodierung geschrieben hat.

Im HTML-Quelltext von b) ist *keine* Zeichenkodierung angegeben. Hier muss der Autor seinen Text in derselben Kodierung abliefern wie der Webserver die Seite ausliefert, an sonsten kommt so etwas heraus wie hier zu sehen.
Wenn allerdings kein oder ein falscher Zeichensatz vom Webserver übermittelt wird fällt das nur auf, wenn der Betrachter, also Du und ich, eine *andere* Zeichenkodierung verwenden als der Autor. Und das ist bei SuSE 9.1 oder 9.2 gar nicht so unwahrscheinlich, weil SuSE seit 9.1 als Zeichenkodierung UTF-8 verwendet, was durchaus sinnvoll ist.
Wenn jetzt der Autor der Webseite seinen Text mit ISO-8859-1 (Linux) oder gar mit Windows verfasst und die Sonderzeichen einfach normal getippt hat, also *keine* HTML-Entities verwendet hat, sehen wir, wenn unser Rechner das Dokument mit UTF-8 anzeigt, nur Blödsinn.
Stell doch mal spaßeshalber die Zeichenkodierung im Konqueror von "Automatisch" auf "Manuell: ISO-8859-1" und alle Sonderzeichen werden korrekt dargestellt. 8-)

Hoffentlich war das verständlich.


Gute Nacht

Pollyanna

Dear Moderator
After kindly reminded that all post should be in English I will extend the original English part of this post so it fitts the whole german text.

Hi,

ein kurzer Nachtrag.

Sowohl mail.yahoo.no, als auch mail.yahoo.de sind auch solche Pappnasen, die keine korrekte Zeichenkodierung angeben, auch nicht im HTML-Dokument.
Die Sonderzeichen werden falsch oder gar nicht dargestellt. Wenn die Seite mal von einem *korrekt* eingestellten Webserver ausgeliefert wird, die haben wohl mehrere davon, kann es ja auch mal mit der Darstellung klappen.


Verärgert

Pollyanna

After kindly reminded that all post should be in English I will translate it.

a short addition.

mail.yahoo.no and mail.yahoo.de are the same kind of clowns, who do not state correct character encoding, not even in the HTML-Document.
Special characters are falsly displayed or not at all. If their website may be delivered by a *working* webserver, I think Yahoo has more than one webserver, the characters may be displayed correctly in your browser.

Angry Pollyanna

Per the LQ Rules "All posts should be in English". Can you please translate your posts? Thanks -- J.W.

both posts were a kind of translation of the english words, so i hope we need no further explanation. but you may be right...
so, in other words, there's nothing a user can do to fix this charset-matter, except from shifting it by hand? a webbrowser should be able to detect it. maybe...

thanks for posting & good night
ungua

Do you mean Firefox? Mozilla's current version is 1.7.3. Firefox is at 1.0, too.

Incidentally I could display both pages with the correct symbols; but in Pollyanna's German message all the characters with diacritical marks are garbled.

Robin