On 2015/08/05 17:46:53, kpreid_google wrote:
> Say "as of 2015-08-05" instead of "at the time of this writing". (The
> information is of course available through history, but giving a
timestamp makes
> staleness instantly understandable.)
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> I suggest swapping these two sentences and putting the new second one
in
> parentheses. This way, the comment begins with what it's actually
about.
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> capitalization: "ECMAScript"
> Do you mean to say it is split _in browsers_ into ...?
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> "URL"
Done.
On 2015/08/04 18:23:35, MarkM wrote:
> delete "three"
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> use <code> not <tt> (here and below).
Done. But why? "<tt>" is much less noisy in the source.
(Makes we wonder whether we should switch to markdown, but of course not
in this CL.)
On 2015/08/04 18:23:34, MarkM wrote:
> delete "so"
Done.
On 2015/08/04 18:23:34, MarkM wrote:
> "mized" --> "mixed"
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> Can be one sentence per line. Rewrap so it is.
Done.
On 2015/08/04 18:23:34, MarkM wrote:
> delete "then"
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> Extract this test into a function returning boolean, so it's less
likely to be
> reinvented or copied, and the comments on how it's a bad test can go
in there
> and not distract from what this is doing.
Done. We were testing whether we're in a browser in two places in
repairES5.js as well, which were using slightly different tests. I
consolidated all these to use the new ses.isInBrowser() test.
In adding ses.isInBrowser to the "// * provides" and "// * requires"
lists at the top of repairES5.js and startSES.js, I noticed various
anomalies there which I also cleaned up.
On 2015/08/04 18:23:34, MarkM wrote:
> Delete "return;". Not needed with this control flow.
Done.
On 2015/08/04 18:23:34, MarkM wrote:
> Delete "Since"
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> Explain why we can't, at least as "See comments on freezeGlobalProp".
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> This will give a false negative if the property's value is undefined.
> This will throw an unhelpful error if the property does not exist
(because desc
> is not an object).
> Parentheses are unnecessary.
Done.
On 2015/08/04 18:23:34, MarkM wrote:
> Always insert a comment in an empty catch clause.
Done.
On 2015/08/04 18:23:34, MarkM wrote:
> Always insert a comment in an empty catch clause.
Done.
On 2015/08/05 17:46:53, kpreid_google wrote:
> Inline this literal into the defProp statement below, so that it can
be read in
> context of it being the descriptor for sharedImports.
Done.
https://codereview.appspot.com/258110043/