[Hours] [Activity] 2011/11/28 Monday 0.25 RT #104504 2.00 reading/responding to list mail 0.25 roffitall ===== 2.50 2011/11/29 Tuesday 1.00 debugger tests 1.75 global destruction backref panic (diagnosis) 0.75 process, scalability, mentoring 1.25 reading/responding to list mail 1.25 use v5.16; ===== 6.00 2011/11/30 Wednesday 2.25 global destruction backref panic (analysis) 5.00 global destruction backref panic (diagnosis) 0.50 process, scalability, mentoring ===== 7.75 2011/12/01 Thursday 1.00 global destruction backref panic (analysis) 2.25 process, scalability, mentoring 2.75 reading/responding to list mail ===== 6.00 2011/12/02 Friday 1.25 process, scalability, mentoring 2.00 reading/responding to list mail ===== 3.25 2011/12/03 Saturday 0.25 checking and merging smoke-me branches 0.50 process, scalability, mentoring 0.25 t/re/regexp.t cleanup ===== 1.00 Which I calculate is 26.50 hours The largest and most "exciting" part of the week was spent single stepping in gdb to determine the cause of a global destruction "panic" related to weak references, that Karl Williamson accidentally stumbled upon whilst innocently refactoring completely unrelated code. After taking the better part of a day to locate and understand the cause, I mailed the explanation to the list, along with some thoughts on what the right fixes are. Several people have provided useful feedback, but I'm hoping that Dave will have time soon to ponder it, as the bug is in the assumptions of changes he made about 8 years ago Nicholas Clark