Apart from looking at a couple of bugs reports, the main thing I did last week was some refactoring of the regex-compiling code which looks for the longest 'must have' fixed and floating substrings in the pattern. Basically I made it so that it maintains a 2-element array of substring info rather than hard-coding for fixed and floating strings. This makes no functional difference at the moment, but opens up the possibility in the future of having more than one of each 'must' substring type. http://nntp.perl.org/group/perl.perl5.porters/245312 2017/06/26 0:38 process p5p mailbox 2:55 fix failing t/op/hash.t test 0:39 refactor compiling regex substring data 2017/06/27 0:20 process p5p mailbox 1:30 [perl #131659] BBC JREFIOR/Finance-StockAccount-0.01.tar.gz 2017/06/29 1:03 process p5p mailbox 1:46 [perl #131660] file:line caller info is changed unexpectedly 7:32 refactor compiling regex substring data 2017/06/30 2:52 refactor compiling regex substring data 2017/07/02 2:00 refactor compiling regex substring data SUMMARY: 1:30 [perl #131659] BBC JREFIOR/Finance-StockAccount-0.01.tar.gz 1:46 [perl #131660] file:line caller info is changed unexpectedly 2:55 fix failing t/op/hash.t test 2:01 process p5p mailbox 13:03 refactor compiling regex substring data ------ 21:15 TOTAL (HH::MM) -- Indomitable in retreat, invincible in advance, insufferable in victory -- Churchill on Montgomery