This allows Qt on Windows to use platform-qt in the inspector.
Created attachment 42482 [details] Patch
I suppose that Gtk and Wx would expect the same behavior.
Looks OK, but I would think better would be to move the WIN_OS option to the end of the list, if it really should be a WIN_OS fallback for platforms based on WIN_OS.
I am not sure what the difference between WIN and WIN_OS is.
(In reply to comment #4) > I am not sure what the difference between WIN and WIN_OS is. WIN_OS is true for any platform building on the Windows operating system. WIN is mostly only true for Apple's Windows port, though it's possible Brent Fulgham's Cairo/libCurl-based Windows port also sets WIN to true.
Oh, you're not alone. :) Sadly our PLATFORM macros are horribly confusing and you might get different answers from different people on different days. :) My understanding is that PLATFORM(WIN) is the Apple windows port, and PLATFORM(WIN_OS) is platforms which sit on top of the win32 api. The story is even more confusing on the mac. Maciej talked about a proposal to fix this long ago, I'm not sure what happened to that.
Created attachment 42516 [details] Alternate solution patch: webkit port identifier If we want to keep a WIN_OS wide platform identifier, I would rather use this solution instead. Unlike the PLATFORM macro on the C++ side, InspectorController.platform() returns a single value that must identify both the webkit flavor and the OS. Tell me if you prefer this solution or the first one, which should use identifers like qt-windows, qt-mac, gtk-linux if needed to switch different behavior/appearance of the inspector on different OS.
Details that might be usefull: currently, the only values returned by platform() that gets used in the code are "mac-*" and "qt"
Comment on attachment 42516 [details] Alternate solution patch: webkit port identifier Clearing flags on attachment: 42516 Committed r50528: <http://trac.webkit.org/changeset/50528>
All reviewed patches have been landed. Closing bug.