FramePolicyFunction should be an std::function
Created attachment 213421 [details] Patch
Comment on attachment 213421 [details] Patch Cute how you made the functions the last argument to match the “block is last argument” style we use for the corresponding Objective-C feature. Not sure if that’s the tradition for C++ but it does seem to make the code readable. I don’t think I’ll ever use a PTMF again!
Comment on attachment 213421 [details] Patch Attachment 213421 [details] did not pass efl-ews (efl): Output: http://webkit-queues.appspot.com/results/3322053
Comment on attachment 213421 [details] Patch Attachment 213421 [details] did not pass win-ews (win): Output: http://webkit-queues.appspot.com/results/3314074
Created attachment 213434 [details] Patch
Created attachment 213436 [details] Patch
Created attachment 213437 [details] Patch
Comment on attachment 213437 [details] Patch Attachment 213437 [details] did not pass win-ews (win): Output: http://webkit-queues.appspot.com/results/3326010
Created attachment 213438 [details] Patch
Committed r156946: <http://trac.webkit.org/changeset/156946>
This is causing link failures on the Mountain Lion bot, probably due to a standard library implementation problem. Various std::function-related symbols seem to be exported from WebKit2 even though we don’t want them to be.
Correction: The problem is on the Lion bot.