Move the tabs to the right one pixel so our tab placement

Move the tabstrip to the right one pixel so our tab placement
matches up with Windows.

BUG=17915

[tony, 2009-07-28 18:43:23]

Here's a before/after/windows screen shot:
  http://ponderer.org/tests/tab_left.png

[James Hawkins, 2009-07-28 18:49:15]

On 2009/07/28 18:43:23, tony wrote:
> Here's a before/after/windows screen shot:
>   http://ponderer.org/tests/tab_left.png

Can you attach a screenshot of "before vs win" and "after vs win".  Also, the
favicon is not lined up with windows anymore.

[James Hawkins, 2009-07-28 18:54:18]

On 2009/07/28 18:49:15, James Hawkins wrote:
> On 2009/07/28 18:43:23, tony wrote:
> > Here's a before/after/windows screen shot:
> >   http://ponderer.org/tests/tab_left.png
> 
> Can you attach a screenshot of "before vs win" and "after vs win".  Also, the
> favicon is not lined up with windows anymore.

Nevermind about the first request.  I'm still concerned about the second issue. 
Why don't we just move the tabs down a pixel instead of shifting them over? 
They will be clipped by the tabstrip and we should have the same affect as
windows.

[tony, 2009-07-28 18:56:59]

On 2009/07/28 18:49:15, James Hawkins wrote:
> On 2009/07/28 18:43:23, tony wrote:
> > Here's a before/after/windows screen shot:
> >   http://ponderer.org/tests/tab_left.png
> 
> Can you attach a screenshot of "before vs win" and "after vs win".  Also, the
> favicon is not lined up with windows anymore.

Here's a copy with guidelines drawn over it:
  http://ponderer.org/tests/tab_left_with_guidelines.png

With this change, the favicons now line up (before, we were off by a pixel).

[Evan Stade, 2009-07-28 19:40:17]

> Why don't we just move the tabs down a pixel instead of shifting them over? 

won't that make them 1 pixel too short

http://codereview.chromium.org/160281/diff/1/2
File chrome/browser/gtk/tabs/tab_strip_gtk.cc (right):

http://codereview.chromium.org/160281/diff/1/2#newcode1365
Line 1365: // shift the tab strip over to account for the difference.
I don't understand this explanation. Vertically we are identical to windows, so
how does the difference of drawing the bottom px vs. not drawing the bottom px
affect the horizontal spacing?

[tony, 2009-07-28 19:49:43]

Opened a bug per jhawkin's request:
http://code.google.com/p/chromium/issues/detail?id=17915

[tony, 2009-07-28 19:49:53]

Yes, my explanation is probably wrong.

[Evan Stade, 2009-07-28 19:55:21]

aside from the explanation, LGTM. The explanation "because it looks better"
seems a good enough reason to me.

[tony, 2009-07-29 00:27:46]

On 2009/07/28 19:55:21, Evan Stade wrote:
> aside from the explanation, LGTM. The explanation "because it looks better"
> seems a good enough reason to me.

Updated the comment to say we're doing it to match Windows.  James, is it ok for
me to check in?  It's needed for matching the rest of the shadow around the
border.

[tony, 2009-07-29 18:10:52]

Updated to offset the whole tabstrip to match windows.  This produces the same
output as before.

James, can I get an LG?

[James Hawkins, 2009-07-29 20:01:53]

On 2009/07/29 18:10:52, tony wrote:
> Updated to offset the whole tabstrip to match windows.  This produces the same
> output as before.
> 
> James, can I get an LG?

Great!  This was what I was looking for.  LGTM.