673104 - Use ES5 strict mode for Panorama

Status: RESOLVED FIXED

(Firefox Graveyard :: Panorama, defect)

Comment 1

[Tim Taubert [:ttaubert] (inactive)]

Attachment: patch v1

Comment 2

[Dão Gottwald [:dao]]

Comment on attachment 547410 [details] [diff] [review]
patch v1

>--- a/browser/base/content/tabview/modules/utils.jsm
>+++ b/browser/base/content/tabview/modules/utils.jsm

>+    function tanh(x) {
>+      var e = Math.exp(x);
>+      return (e - 1/e) / (e + 1/e);
>+    }
>+
>     if (smooth) {
>       // The ease function ".5+.5*Math.tanh(4*x-2)" is a pretty
>       // little graph. It goes from near 0 at x=0 to near 1 at x=1
>       // smoothly and beautifully.
>       // http://www.wolframalpha.com/input/?i=.5+%2B+.5+*+tanh%28%284+*+x%29+-+2%29
>-      function tanh(x) {
>-        var e = Math.exp(x);
>-        return (e - 1/e) / (e + 1/e);
>-      }
>       return .5 - .5 * tanh(2 - 4 * proportion);
>     }

>--- a/browser/base/content/tabview/search.js
>+++ b/browser/base/content/tabview/search.js

>+  function dispatchTabViewSearchEnabledEvent() {
>+    let newEvent = document.createEvent("Events");
>+    newEvent.initEvent("tabviewsearchenabled", false, false);
>+    dispatchEvent(newEvent);
>+  }
>+
>   if (!isSearchEnabled()) {
>     $searchShade.show();
>     $search.show();
> 
> #ifdef XP_MACOSX
>     UI.setTitlebarColors({active: "#717171", inactive: "#EDEDED"});
> #endif
> 
>     // NOTE: when this function is called by keydown handler, next keypress
>     // event or composition events of IME will be fired on the focused editor.
>-
>-    function dispatchTabViewSearchEnabledEvent() {
>-      let newEvent = document.createEvent("Events");
>-      newEvent.initEvent("tabviewsearchenabled", false, false);
>-      dispatchEvent(newEvent);
>-    };

That's annoying. What rule enforces this?

Comment 3

[Tim Taubert [:ttaubert] (inactive)]

This one:

http://whereswalden.com/2011/01/24/new-es5-strict-mode-requirement-function-statements-not-at-top-level-of-a-program-or-function-are-prohibited/

Comment 4

[Dão Gottwald [:dao]]

Again, annoying. Does let foo = function ... work?

Comment 5

[Tim Taubert [:ttaubert] (inactive)]

Attachment: patch v2

(In reply to comment #4)
> Again, annoying. Does let foo = function ... work?

Yep, it does.

Comment 6

[Dão Gottwald [:dao]]

Comment on attachment 547523 [details] [diff] [review]
patch v2

> function observer(subject, topic, data) {
>   switch (topic) {
>     case "domwindowopened":
>-      subject.addEventListener("load", function() {
>-        subject.removeEventListener("load", arguments.callee, false);
>+      subject.addEventListener("load", function onLoad() {
>+        subject.removeEventListener("load", onLoad, false);

I wonder where this makes onLoad available, exactly, and why this doesn't bother strict mode.

Comment 7

[Tim Taubert [:ttaubert] (inactive)]

(In reply to comment #6)
> I wonder where this makes onLoad available, exactly, and why this doesn't
> bother strict mode.

In this case it's a function expression (not a function definition) and so the name of it is bound locally inside the function, but not outside.

Comment 8

[Tim Taubert [:ttaubert] (inactive)]

http://hg.mozilla.org/integration/fx-team/rev/f79f1ea87e2c

Comment 9

[Tim Taubert [:ttaubert] (inactive)]

http://hg.mozilla.org/mozilla-central/rev/f79f1ea87e2c