네트워크의 진화

Note: Expanded support has been added to vB 4.0.2.

Three new options exist for controlling Friendly URL behaviour. The suitable values will differ from board to board depending on wether multiple character sets are / have been in use and on the admin's preference:

Friendly URL Unicode
Ignore / Convert to UTF-8 / Strip Unicode

Determines how to deal with multiple character sets in URLs. If you use a single character set and most of your content is submitted as that character set then Ignore or Strip may be the best options. Ignore attempts to form the URL from the raw content, while strip will remove unicode encoded characters.

Convert to UTF-8 will convert the content to UTF-8 from the current language characterset and is likely to be the best solution if a lot of your content is submitted in a different character set than the language (such as submitting arabic on an ISO-8859-1 page).

Enforce Canonical URL
Off / Standard / Strict

Standard only checks the Friendly URL method used for the request. If it does not match the board setting then the user will be redirected to the correct URL. If Strict is selected then the URL content will also be checked. In some cases this may be tripped up by differing encoding, but provides a way for admins to make the check less strict if a problem occurs.

In addition, strict canonical checking will try both UTF-8 and the language charset. Canonical URLs are now only enforced on guests.

NCR Encode UTF-8 URLs for IE
Yes / No
If selected then Friendly URLs will be NCR encoded if the user is using Internet Explorer. This is more likely to guarantee that URLs are displayed correctly but requires a little overhead and larger output.

Note: Internet Explorer has no 'correct' method for outputting URLs in a character set that differs from the page. The W3 standard is to send URLs as urlencoded UTF-8, however IE will display this raw (ie showthread.php?1-%5D%3E%FA etc) while all other browsers convert them to be displayed as characters.