Issue 1197 - adapt gyp variables

Define v8_use_external_startup_data gyp variable, without it libadblockplus\third_party\v8\tools\gyp\v8.gyp (from v8 4.3.15) fails to generate project files.

For additional details see v8 commit 68870ab6ee57a92d140dd2c26b0639fbe84975a6.
In particular, in that commit v8_use_external_startup_data has been moved to standalone.gypi. We don't use standalone.gypi so we have to manually define it now.
As well as using that commit one can find the background of the change using bug ID (https://code.google.com/p/chromium/issues/detail?id=421063) and review URL.

[sergei, 2016-01-25 15:36:43]

Update reviewers.

[Oleksandr, 2016-01-26 05:30:20]

Just in case here is some more context for the v8 change:
https://code.google.com/p/chromium/issues/detail?id=421063

LGTM

[Eric, 2016-01-26 18:10:43]

Commit message needs better documentation about what this change is doing.
Pointing to a particular git commit ID is mostly useless to understand what's
going on here. The URL Oleksandr pointed is much better reference information.
Nevertheless, the commit message still needs to self-contained documentation
here.

[sergei, 2016-01-27 09:23:15]

On 2016/01/26 18:10:43, Eric wrote:
> Commit message needs better documentation about what this change is doing.
> Pointing to a particular git commit ID is mostly useless to understand what's
> going on here. The URL Oleksandr pointed is much better reference information.
> Nevertheless, the commit message still needs to self-contained documentation
> here.

I have updated the description.
However, I disagree that pointing to a particular git commit is useless because
the change which breaks our project has been done in that particular commit and
one can find out what happened using the commit message and the diff as well as
if somebody is interested in the background he can find the bug ID and review
URL in the commit message. On the bug page there are really many links to
commits and reviews and the information is obscured by a long flow of
discussion.

[Eric, 2016-01-27 15:15:21]

On 2016/01/27 09:23:15, sergei wrote:
> I have updated the description.

The description is adequate now. It explains why there's a new line in the gyp
file.

> However, I disagree that pointing to a particular git commit is useless 

I did not say "useles". I said "mostly useless". You don't need to take it out.
But you should move it to the bottom of the commit message, as it acts solely as
reference information.

Let me be clearer. We need, in this commit message, the version number of v8 in
which the referenced change set was first published as RTM code.

[sergei, 2016-01-29 12:59:20]

On 2016/01/27 15:15:21, Eric wrote:
> On 2016/01/27 09:23:15, sergei wrote:
> > I have updated the description.
> 
> The description is adequate now. It explains why there's a new line in the gyp
> file.
> 
> > However, I disagree that pointing to a particular git commit is useless 
> 
> I did not say "useles". I said "mostly useless". You don't need to take it
out.
> But you should move it to the bottom of the commit message, as it acts solely
as
> reference information.
> 
> Let me be clearer. We need, in this commit message, the version number of v8
in
> which the referenced change set was first published as RTM code.

Version number of v8 is defined in the issue. This change is rather a "no issue
commit" for the branch issue-1197 but we don't use branches. Instead on practice
we use labels "issue xxx" in the prefix of a commit title. In addition, it's not
in https://codereview.adblockplus.org/6193234183192576/ because it is a
self-contained change and it reliefs that codereview (6193234183192576).

[Eric, 2016-02-09 15:06:27]

On 2016/01/29 12:59:20, sergei wrote:
> Version number of v8 is defined in the issue. 

I see "Issue 5145227803230208" and I don't see a v8 version number anywhere on
this page.

If you're talking about a ticket in Trac, that's not what I'm asking for. We
need a version number of v8 in the commit message for the present change set.

> In addition, it's not
> in https://codereview.adblockplus.org/6193234183192576/ because it is a
> self-contained change and it reliefs that codereview (6193234183192576).

It's fine that it's self contained. It still needs a version number of v8 of its
own so that it can be audited against an externally-changing v8 code base.

[sergei, 2016-02-18 10:52:28]

On 2016/02/09 15:06:27, Eric wrote:
> On 2016/01/29 12:59:20, sergei wrote:
> > Version number of v8 is defined in the issue. 
> 
> I see "Issue 5145227803230208" and I don't see a v8 version number anywhere on
> this page.
> 
> If you're talking about a ticket in Trac, that's not what I'm asking for. We
> need a version number of v8 in the commit message for the present change set.
> 
> > In addition, it's not
> > in https://codereview.adblockplus.org/6193234183192576/ because it is a
> > self-contained change and it reliefs that codereview (6193234183192576).
> 
> It's fine that it's self contained. It still needs a version number of v8 of
its
> own so that it can be audited against an externally-changing v8 code base.
Added `(from v8 4.3.15)` into the commit message.