Update status file to fix tests that are now passing because of FF on windows update.

tests/html/html.status

 # Copyright (c) 2012, the Dart project authors.  Please see the AUTHORS file
 # for details. All rights reserved. Use of this source code is governed by a
 # BSD-style license that can be found in the LICENSE file.
 
 event_test: Skip  # Issue 1996
 webgl_1_test: Skip  # Issue 1495
 xmldocument_test: Skip  # XML in dart:html is incomplete
 xmlelement_test: Skip  # XML in dart:html is incomplete
 measurement_test: Pass, Fail  # Issue 1940
 
@@ skipping 120 matching lines @@
 js_interop_4_test: Fail
 js_interop_func_passing_test: Fail
 js_interop_obj_passing_test: Fail
 js_interop_obj_invoke_test: Fail
 js_interop_element_test: Fail
 # Interfaces not implemented: SVGTests, SVGLangSpace, SVGExternalResourcesRequired, SVGStylable
 svg_3_test: Fail
 svgelement_test: Fail
 svgelement2_test: Fail
 transferables_test: Fail   # Issue 3392.
-typed_arrays_dataview_test: Fail  # DataView not implemented until Firefox 15.0
 # document.getElementsByTagName('body') returns an HTMLCollection, not a NodeList.
 #   TODO: Make both implement List<Node>, make implementation return List<Node>, not _NodeListJs.
 # Note this test now passes for the wrong reasons. It used to be that the only
 # forEach in this program was in NodeList. Due to dependencies from the unittest
 # framework, frog no longer resolves forEach statically, and instead it
 # dynamically binds forEach on a bunch of classes, including HTMLCollection.
 typing_test: Pass
 websql_test: Fail # FF does not support web SQL
 websocket_test: Pass, Fail # FF 10 is uses MozWebSocket, FF 11 uses WebSocket.
 # Exception... "Access to restricted URI denied" code: "1012" nsresult: "0x805303f4
 #    (NS_ERROR_DOM_BAD_URI) ...
 xhr_test: Fail
 
+[ $runtime == ff && $system == linux ]
+typed_arrays_dataview_test: Fail  # DataView not implemented until Firefox 15.0. Currently only the windows bot is running FF 15 - see issue 4896.

[kasperl, 2012/09/04 07:08:05]

Maybe make the comment fit better on one line?

[ngeoffray, 2012/09/04 07:14:51]

Do we expect ff tests to run only on bots? Maybe better mark this test as Fail,
Pass to not get conflicting results when running the tests locally?

[kasperl, 2012/09/04 07:17:34]

That's a good point -- but I think we need some kind of general way of dealing
with this. It will only get worse over time. I guess we should be able to have
the separate expectations for different browser versions -- and maybe have
checks that ensure that we're not trying to test on too old browsers (to avoid
keeping the expectations up-to-date for very old browsers).

[ricow1, 2012/09/04 07:23:33]

Done.

[ricow1, 2012/09/04 07:23:33]

I really don't like tests marked Fail, Pass - this is essentially the same as
marking the test as Skip (except that you _will_ get crashes)
FF 15 is current stable release of FF afaik, and FF already have a setup similar
to chrome in that it will do automatic updates of users (unless disabled)

+
 [ $runtime == ie && ($system == linux || $system == macos) ]
 *: Skip
 
 [ $runtime == safari && ($system == linux || $system == windows) ]
 *: Skip
 
 [ $runtime == safari ]
 # TODO(vsm): Triage DOM failures.
 *: Skip
 
@@ skipping 41 matching lines @@
 xhr_test: Skip # Timeout
 inner_frame_test: Skip # Timeout
 svg_3_test: Skip # Timeout
 documentfragment_test: Skip # Timeout
 
 [ $compiler == dart2js && $runtime == chrome && $system == windows]
 css_test: Pass, Fail # Issue #2823
 
 [ $compiler == dart2js && $runtime == ff && $system == windows]
 js_interop_2_test: Fail # Issue 4658.