default cfm/file charset problem source found

[richso]

the default cfm/file charset problem source has been found, it's SUN's java
development teams decision for standardize the system setting of
'file.encoding' to be readonly and follow the underlying OS setting only. many
people has found the setting can be overriden in some system environment but
not others, so do try it by specify -Dfile.encoding=your_encoding, if this work
then you are lucky; however, it's not guarantee work for another new version of
jvm >_<. reference: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=4163515
failed to use the setting, you may need to use <cfprocessingdirective
pageencoding='your_encoding'> on EVERY file and specify encode='your_encoding'
in file processing tags !! >_<

[richso]

this is actually a very bad situation for the guys, like me, who are going to
migrate to use unicode but in the mean time use the old encoding. Because we
need to make the system reversible to old version in case for some problems
found cannot be easily tackled in new version, we will need to enable the code
runs on the old version and MX; however this CANNOT BE DONE now because we must
specify <cfprocessingdirective pageencoding='your_encoding'> for every files
and the tag will cause a just-in-time compiler error in version 4.5 (I don't
know for 5.0) and cannot be <cftry> trapped. SO, PLEASE DO YOUR OBJECTION IN
JAVA.SUN.COM IF YOU FOUND THE SETTING SHOULD BE OVERRIDE-ABLE !!

[PaulH]

you might try changing the defaultCharset value in cf_root/lib/neo-runtime.xml
file:

<var name='defaultCharset'><string>UTF-8</string></var>

from UTF-8 to whatever you require (and core java supports).