I'm transferring my application from a tomcat 5.5.26 server to tomcat
6.0.18, and notice that my formatted currency amounts are not being properly
displayed. Instead of a Pound (GBP) sign I get a question mark within a
black diamond (the app works fine in 5.5.26).
This can easily be emulated. Add
Works fine for me, fresh install of 6.0.18, changed the
HelloWorldExample.java and
recompiled.
Tried with both IE7 and FF 3.
Are you sure you don't have a httpd in front of tomcat?
I've seen simillar problem when using apache httpd.
I had to turn off the option
AddDefaultCharset
On Wed, Sep 10, 2008 at 7:55 PM, Steve Ochani [EMAIL PROTECTED] wrote:
Hmm odd.
I tried it on my Redhat test server and worked fine also.
Is your tomcat 6 install a default/fresh install?
What browser are you using? What character encoding does it think the
HelloWorldExample
output is
Will if possible use
pound
instead... that I think its font independent...
Otherwise I think you have to sorround that
getCurrencyInstance
stuff with a font... and tell it what font it must use...
... I think
I'm just wondering how the systems guess the character set from
1. What the _Browser_ thinks about encoding of your page.
In menu View Encoding what encoding is auto-selected there.
Western / ISO 8859-1 for both.
2. In Page Info dialog of Firefox
(in Tools menu or in context menu Page Info )
what is Encoding, Content Type, and what META tags are
On Thu, Sep 11, 2008 at 11:35 AM, Mark Hagger [EMAIL PROTECTED]wrote:
You are almost certainly having a problem with (default) character
encodings on your system, usual things to check are the encoding that
the JVM is using, for example what does:
echo $LANG
return (usually controlled by
On Thu, Sep 11, 2008 at 1:42 PM, Johnny Kewl [EMAIL PROTECTED] wrote:
Will, I cant see how TC can be influencing it
You write a char (the pound) to an output stream it appears differently in
browser...
TC is just sendign what it gets...
Its got to be this...
I studied the Response Headers for the ajax call that generates the output
and found that for the correct result (ie. in TC55), the content type was
this:
Content-Typetext/plain;charset=ISO-8859-1
while for the wrong result (ie. in TC6), the content type was:
Content-Typetext/plain
So I
On Fri, Sep 12, 2008 at 9:26 PM, Johnny Kewl [EMAIL PROTECTED] wrote:
Wonder if Wil knew he asked such a damn big question... ha ha
I'm really amazed at the volume of mails my question has raised.
I can only see one solution to this complexity: let's all (everybody in the
whole world) speak