powderkeg

viesti 2017-03-31T11:41:26.863747Z

hmm, master...

viesti 2017-03-31T11:41:45.866270Z

maybe into sql branch first

viesti 2017-03-31T11:46:36.905744Z

conflict in project.clj but not too bad

cgrand 2017-03-31T11:47:31.912827Z

I like when you call me β€œmaster” πŸ˜‰

viesti 2017-03-31T11:52:02.950183Z

πŸ˜„

viesti 2017-03-31T11:53:03.958380Z

a neat welcome to pop off on the cellphone πŸ™‚

cgrand 2017-03-31T13:28:01.004857Z

this ::mapping spec is monstruous

viesti 2017-03-31T13:29:26.024945Z

πŸ˜„

viesti 2017-03-31T13:29:57.032410Z

could it be expressed without a spec at all?

cgrand 2017-03-31T13:30:45.044577Z

its main problem to me is that it’s doing too many things at once

viesti 2017-03-31T13:57:22.444887Z

as in something could be deferred to be done later or too much code inline?

cgrand 2017-03-31T14:01:31.514419Z

given a spec, we want: β€’ a schema β€’ converting to a row β€’ the serializer part of the encoder β€’ converting from a row β€’ the deserializer part of the encoder

cgrand 2017-03-31T14:02:08.524477Z

given a schema, we want a spec (and from this spec gets the conversion fns)

cgrand 2017-03-31T14:02:51.536857Z

::mapping tries to do everything at once (modulo encoders)

viesti 2017-03-31T19:44:54.942369Z

hmm, master...

viesti 2017-03-31T19:45:41.953202Z

is ::mapping currently to much to be fed into master branch?

viesti 2017-03-31T19:46:01.957544Z

just had to try out the β€œmaster” summon again πŸ™‚

cgrand 2017-03-31T19:46:41.966760Z

yes too hacky even by powderkeg standards

viesti 2017-03-31T19:47:11.973133Z

πŸ˜„