Remove misleading comment for $wgLegalTitleChars

This comment originates from rSVN1420 (9d51f616), dated 2 July 2003,
where it was written as "ISO 8859-* don't allow 0x80-0x9F... But that
breaks interlanguage links at the moment". It was rephrased to the
current form in rSVN2621 (840dee3a).

It is incorrect for two reasons:

* "Theoretically 0x80-0x9F of ISO 8859-1 should be disallowed..."

  We cannot disallow 0x80-0x9F here; this config variable actually
  specifies the valid ranges of *bytes* rather than characters, and
  0x80 to 0x9F can happily appear in valid UTF-8 encodings of other
  characters.

  In case we wanted to disallow the Unicode characters U+0080 to U+009F
  (encoded in UTF-8 as 0xC2 0x80 to 0xC2 0x9F), it would probably have
  to be done explicitly in MediaWikiTitleCodec::splitTitleString().
  (The task for this is T7732.)

* "...but this breaks interlanguage links"

  Back then, most wikis were using single-byte ISO encodings rather
  than UTF-8, and that is the only configuration this comment applies
  to: disallowing the bytes 0x80-0x9F in page titles on wikis using
  single-byte ISO encodings would indeed have broken interlanguage
  links from them to wikis using UTF-8. However, disallowing the
  Unicode characters U+0080 to U+009F today definitely would not break
  interlanguage links.

Change-Id: Ic5ba502ccfbb9cf3ff56cc47eb7fe463e7d45959

diff --git a/includes/DefaultSettings.php b/includes/DefaultSettings.php
index 8f4c346..2b2695c 100644 (file)
--- a/includes/DefaultSettings.php
+++ b/includes/DefaultSettings.php
@@ -3936,9 +3936,6 @@ $wgNamespaceAliases = [];
  * because articles can be created such that they are hard to view or edit.
  *
  * In some rare cases you may wish to remove + for compatibility with old links.
- *
- * Theoretically 0x80-0x9F of ISO 8859-1 should be disallowed, but
- * this breaks interlanguage links
  */
 $wgLegalTitleChars = " %!\"$&'()*,\\-.\\/0-9:;=?@A-Z\\\\^_`a-z~\\x80-\\xFF+";