FS Always mark xml as ISO-8859-1, but pass info as-is.
So if user send its contact info as UTF-8 then we get
corrupted xml. But RFC describe text as UTF8
RFC 3261 - 25.1 Basic Rules
```
qdtext = LWS / %x21 / %x23-5B / %x5D-7E / UTF8-NONASCII
```
convert more messages to messages class
skipped dashboard messages due to them relying on $_SESSION['message']
to know if there has already been a change
make spacing and layout consistent
add all missing language tags
fill missing en-us descriptions
convert to $x$y$z usage for array filling
add a default version number
this is a by hand update
Imported Found language ro as ro-ro Romanian (tested by reverse translation)
Add placeholders for all found languages
Add new missing flags
Enhancements to text class
* improved sort to keep language-* tags organized at the top
* if a language is missing a name tag, add it
* changed organizer to not add a final \n
* corrected some indenting
* added detect_all_languages that attempts to find any languages that haven't been defined in resources/app_languages.php
* added language_totals, this is used by the optional languages application to calculate how complete a language is
convert all app_menu.php files to have all placeholders and languages and use 2-2 format for languages
update app_defaults to convert any items found using legacy values in user_setings, domain_settings and default_settings
Convert all languages to use the en-us style codes
Where an installation is using a 2 digit code, use the legacy map to find
the mapped language and provide that text correctly
Have escape_str strip \' before converting or we end up with a double
encode
Update organzie_languages to understand how to convert legacy languages
automatically
updated flags
organize_languages will now use an internal function to escape values as fusion's check_str would not be appropriate
enabled two missing flags
organize on remaining app_languages.php and add missing 'he'
