Fix for IE6 FRP crash

Fix for IE6 FRP crash

[Oleksandr, 2013-08-11 06:11:08]

This fixes 2 issues with FRP in fact. 

1. Crashing
2. Actual loading of data from the plugin.

[Felix Dahlke, 2013-08-11 11:23:12]

http://codereview.adblockplus.org/11364147/diff/1/html/templates/firstRun.html
File html/templates/firstRun.html (right):

http://codereview.adblockplus.org/11364147/diff/1/html/templates/firstRun.htm...
html/templates/firstRun.html:71: <script
src="../static/js/firstRun.js"></script>
I don't think moving the scripts here is necessary, the init code is only
running after window's load event is emitted, isn't it?

http://codereview.adblockplus.org/11364147/diff/1/src/plugin/PluginClass.cpp
File src/plugin/PluginClass.cpp (right):

http://codereview.adblockplus.org/11364147/diff/1/src/plugin/PluginClass.cpp#...
src/plugin/PluginClass.cpp:816: // IE6 can't be accessed from another thread,
execute in current thread
So you mean we cannot create threads in IE? If so, could you put it that way in
the comment? Also, won't that bite us elsewhere?

[Oleksandr, 2013-08-11 16:26:53]

http://codereview.adblockplus.org/11364147/diff/1/src/plugin/PluginClass.cpp
File src/plugin/PluginClass.cpp (right):

http://codereview.adblockplus.org/11364147/diff/1/src/plugin/PluginClass.cpp#...
src/plugin/PluginClass.cpp:816: // IE6 can't be accessed from another thread,
execute in current thread
What I mean is that IE6 COM object can not be controlled by our threads. It
looks like those threads have some other initialization besides CoInitialize. I
think my comment describes it more precise. Do you insist on changing it?

No, I don't think there are any problems with this elsewhere. In all other cases
we use IE in it's threads.  
On 2013/08/11 11:23:12, Felix H. Dahlke wrote:
> So you mean we cannot create threads in IE? If so, could you put it that way
in
> the comment? Also, won't that bite us elsewhere?

[Felix Dahlke, 2013-08-11 16:58:18]

LGTM

http://codereview.adblockplus.org/11364147/diff/1/src/plugin/PluginClass.cpp
File src/plugin/PluginClass.cpp (right):

http://codereview.adblockplus.org/11364147/diff/1/src/plugin/PluginClass.cpp#...
src/plugin/PluginClass.cpp:816: // IE6 can't be accessed from another thread,
execute in current thread
On 2013/08/11 16:26:53, Oleksandr wrote:
> What I mean is that IE6 COM object can not be controlled by our threads. It
> looks like those threads have some other initialization besides CoInitialize.
I
> think my comment describes it more precise. Do you insist on changing it?
> 
> No, I don't think there are any problems with this elsewhere. In all other
cases
> we use IE in it's threads.  
> On 2013/08/11 11:23:12, Felix H. Dahlke wrote:
> > So you mean we cannot create threads in IE? If so, could you put it that way
> in
> > the comment? Also, won't that bite us elsewhere?
> 

No, I think the current one nails it then.

[Wladimir Palant, 2013-08-13 09:52:31]

LGTM (see nit below that needs addressing)

http://codereview.adblockplus.org/11364147/diff/1/html/static/js/ieFirstRun.js
File html/static/js/ieFirstRun.js (right):

http://codereview.adblockplus.org/11364147/diff/1/html/static/js/ieFirstRun.j...
html/static/js/ieFirstRun.js:46: return window.Settings.GetMessage(section,
param);
Tabs used instead of spaces for indentation in this file?