salvaged (e.g., need locale info to properly predict whether we're subject to
DST).

recoded a file to ANSI).

need to move some tests over to best practices.

All Microsoft-contributed *.js and *.py sources now have Ecma's copyright header.

https://bugs.ecmascript.org/show_bug.cgi?id=266 and https://bugs.ecmascript.org/show_bug.cgi?id=270 resolved.

Fixed.

Re-gen'ed the website.

Extension clause permits function declarations more or less anywhere.
Moved four such negative test cases to best practices directory.

- https://bugs.ecmascript.org/show_bug.cgi?id=44
- https://bugs.ecmascript.org/show_bug.cgi?id=45

all 'N'.  If such files existed, they've been moved to test/suite/chN/N.0/
now.

architectures), but worth keeping around for the coverage it provides.
Documented that this is expected to fail on ARM.

S7.8.5_A3.1_T7 through S7.8.5_A3.1_T9 were invalid as written.  The only way to specify unicode flags
to literal regexp patterns is through eval and similar mechanisms (e.g., Function constructor).  Fixed.

All but two of the remaining disabled tests in excludelist.xml were disabled because the tests polluted
the global JS environment, not because the tests themselves were invalid.  This is no longer an issue as
of the latest test262 harness (actually it wasn't an issue about a year ago either).  Re-enabled.

Mark Miller removed this invalid test in Sputnik (see http://code.google.com/p/sputniktests/source/detail?r=93), and
I'm propagating that change to test262 now.

Also, two of this weren't actually this-related.  Fixed their separate issues.

S11.1.1_A2.js is invalid and there's no way to make it valid.  Removing.