05:32:45 <mikeplus32> any plans/progress in migrating from [Char] to Text/ByteString (for instance, in cookies)? the main reason i dno't use happstack is because of semi-paranoid - but grounded - concerns about performance such as its use of [Char]
06:59:38 <stepcut> mikeplus32: yes -- working on a new backend that uses pipes/bytestring/Text
06:59:53 <stepcut> mikeplus32: with a test suite that includes performance benchmarks
07:00:24 <mikeplus32> stepcut: great, any repo available that i can contribute to?
07:00:37 <stepcut> mikeplus32: happstack pre-dates ByteString/Text -- so some parts are still based on String -- but working on replacing those parts
07:01:23 <stepcut> mikeplus32: in theory, here, https://github.com/stepcut/hyperdrive, but that is actually out of date
07:01:47 <stepcut> mikeplus32: practically speaking, it is still a tad early to contribute -- still hashing out the very fundamental parts
07:02:03 <stepcut> and, things like pipes-parse are still going though design changes
07:02:56 <mikeplus32> alright
07:05:52 <stepcut> the old backend works pretty darn good -- so spending the time needed to get the new backend right since there is no urgent need to get it out fast
07:07:12 <mikeplus32> fair enough, i suppose i'll continue using snap, until this new backend gets worked out (i much prefer happstack's design, generally)
07:08:43 <stepcut> sounds good -- I am confident that any of the haskell web frameworks will get the job done :)
08:11:40 <mm_freak> stepcut: that reminds me…  is there a reason you don't use http-types?
16:19:54 <stepcut> mm_freak_: nope -- plan to switch unless something comes up
16:39:18 <donri> it does seem to follow the spec nicely, from the limited time i've spent looking at it
16:59:57 <stepcut> the bigger issue at the moment is how to allow the user supplied handler to consume the request body
17:00:08 <stepcut> preferably with out relying on IORef