Bug 1563278: Specify font-family and font-size on the same element in web-platform-test contain-paint-008.html. r=TYLin

This patch corrects a likely-inadvertent difference between the CSS in this
testcase vs. its reference case.

This patch fixes a Windows & MacOS off-by-1px spurious failure in this test,
which was probably due to having text whose font-metrics differ from its
line-box (in the testcase) vs. agree with its line-box (in the reference).

While we're here, this patch also adjusts the code-comment in the testcase
to avoid implying that Firefox unconditionally lacks support for containment.
(And to fix a minor grammatical typo, "not need" --> "no need".)

Differential Revision: https://phabricator.services.mozilla.com/D37494

--HG--
extra : moz-landing-system : lando

testing/web-platform/meta/css/css-contain/contain-paint-008.html.ini

@@ -1,5 +0,0 @@
-[contain-paint-008.html]
-  expected:
-    if (os == "win"): FAIL
-    if (os == "mac"): FAIL
-  bug: https://bugzilla.mozilla.org/show_bug.cgi?id=1563278

testing/web-platform/tests/css/css-contain/contain-paint-008.html

@@ -13,18 +13,18 @@ rt {
   contain: paint;
   display: ruby-text;
   font-size: 1rem;
+  font-family: monospace;
 }
 rt::after {
   content: "PASS";
 
-  /* Doing the following instead of position:absolute to move it out into the area that would be clipped
-     because Firefox clips absolutely positioned content of rt even though it does not support
-     containment.
-     Since this technique works also, not need to trigger a false negative.
+  /* Doing the following instead of position:absolute to move it out into the area that would be
+     clipped because Firefox clips absolutely positioned content of rt even in builds without
+     support for css containment.
+     Since this technique works also, there's no need to trigger a false negative.
    */
   position: relative;
   left: 4ch;
-  font-family: monospace;
 }
 </style>