00:00:25 <mm_freak> i said resource because i thought of this typical usage:  x `finally` hClose h
00:00:41 <mm_freak> you don't want hClose h to be run before x is actually finished
00:02:18 <stepkut> hmm
01:33:34 <stepkut> interesting
01:35:51 <stepkut> its not so much that the morphism allows you to escape the whole resource management as it is trySafeIO blowing the head of the thing and causing completely unpredictable behavior
01:36:24 <stepkut> but.. possibly just an implementation bug
01:47:15 <stepkut> i think it can be prevented with a newtype
02:04:42 <stepkut> or not
02:04:44 <stepkut> getting too tired
02:15:18 <stepkut> I emailed tekmo
02:15:54 <hpaste> “trySafeIO pipes-safe explosion” pasted “stepcut” at http://hpaste.org/87219
03:47:57 <stepkut> apparently that is fixed in pipes-safe-2.0, which also no longer requires the SafeIO monad either :)
05:08:09 <mm_freak> stepkut: did he already answer?
12:46:13 <stepkut> mm_freak: yup
13:06:48 <mm_freak> stepkut: may i reply to the mail?
20:10:48 <mreh> stepcut: I want to help with clckwrks