Clojurians Log v2
Clojure programming
Channels
Apps
malli
https://github.com/metosin/malli :malli:
ikitommi 2020-06-12T06:10:01.032300Z
2020-06-12T17:18:50.042100Z
@ikitommi My concern about the size was from the perspective of delivering a very lightweight web framework (Vrac) which can compare well with others in a benchmark. I acknowledge that, in real world applications, the size difference is not as important as it may look on a benchmark.
2020-06-12T17:19:29.042300Z
Malli’s size is still relatively small.
ikitommi 2020-06-12T06:10:34.033Z
initial bundle-size anaysis for cljs. related to #203
borkdude 2020-06-12T09:36:50.033300Z
@ikitommi Note that about 200kb of the sci bundle is coming from docstrings 🙂
borkdude 2020-06-12T09:37:02.033600Z
at least, if you bump to the newest sci
ikitommi 2020-06-12T10:07:04.038400Z
yes, updated the latest. Haven’t really studied how the dce works and how can libraries be smaller. Would be great if one could drop off whole protocol method impls if the protocol method was not used. E.g. if you don’t have any calls to the transformers, we coud remote the -transformer methods, which is easily 30% of the whole malli.core.
ikitommi 2020-06-12T10:07:52.039400Z
most of the concerns are in separate namespaces (json-schma, human errors, inferring), but the important ones are in the core right now. But not always important 🙂
ikitommi 2020-06-12T10:08:25.040100Z
also, would be great to have a sci-liteversion without the docstrings.
borkdude 2020-06-12T10:09:14.040900Z
maybe that can be accomplished using a goog/define
borkdude 2020-06-12T10:10:08.041400Z
:closure-defines {"sci.foo/DOCSTRINGS" false}borkdude 2020-06-12T10:11:12.041600Z
made an issue for it: https://github.com/borkdude/sci/issues/352