It seems that a lot of people don’t know how cmd click works in a lot of Mac OS X Cocoa applications. To put it simply, Cmd-clicking on a non key window (or application for that matter) will act like a normal click without making the window key! This is really cool for testing UI things while debugging or testing for memory leaks.<\/p>\n
Here is how I use this technique while debugging:
\n1. I’m debugging my application with Xcode. Inside of Xcode I have a \u00e2\u20ac\u0153hot\u00e2\u20ac\u009d breakpoint that I don’t want to enable, since enabling it will make Xcode key, and the key\/focus switching will hit the breakpoint at the wrong time.
\n2. Because of this, I keep my application key, and Cmd-click on the title bar in Xcode to move the window to where I can see the breakpoint.
\n3. Then, I Cmd-click the breakpoint, enabling it without setting the window key and never leaving focus from my application that i want to debug.
\n4. Now, I perform the operation that invokes the breakpoint and debug away.<\/p>\n
Sure, you could do this by figuring out how many times the breakpoint was hit, and in gdb ignore the breakpoint for X times, but sometimes that is a pain to do.<\/p>\n
I also use this technique while trying to find memory leaks with Object Alloc. Frequently, focus switching causes numerous allocations, and by Cmd-clicking the \u00e2\u20ac\u0153Mark\u00e2\u20ac\u009d button in Object Alloc I can prevent those allocations from happening.<\/p>\n
Note that cmd click doesn’t work 100% correctly with certain components. For instance, NSTableView won’t change the selected item unless it is key (note that that bug will be fixed).<\/p>\n
–corbin<\/p>\n","protected":false},"excerpt":{"rendered":"