Fork me on GitHub
#announcements
<
2019-06-05
>
seancorfield01:06:21

seancorfield/next.jdbc "1.0.0-rc1" -- fixes reflection warnings from get-connection; adds more flexibility to the result set builders, including a new next.jdbc.optional namespace providing result set map builders that omit columns with null values, per "Optionality" discussions. All follow-up to #sql -- docs https://cljdoc.org/d/seancorfield/next.jdbc/1.0.0-rc1/

tony.kay06:06:40

Fulcro 3 is well on its way to being production ready. Version 3.0.0-alpha-2 is on Clojars, and the README has been updated to explain how this new version is improving. Fulcro 3 uses all new namespaces. It is “nearly” source compatible with Fulcro 2, but many of the simplifications and improvements change a number of things that merited new names, and a formal porting step. This is living in an alternate repository for now, but will probably be folded back to the original when I’m done: https://github.com/fulcrologic/fulcro3

mattias50406:06:51

Interesting! Being new in the ecosystem, I’d be very interested in a reflection on how Fulcro should be considered compared to alternatives - if there are any. I have found that type of information a bit hard to find, at times, so any pointer would be great. And yes, ideally I would implement something to know for myself, but, we’ll... 😅

jacek.schae07:06:24

https://clojurescriptpodcast.com/ ClojureScript Compiler with Mike Fikes

pez14:06:42

I have just published a significant update of Calva, sometimes referred to as Calva 2 (or Calva dos maybe a Spanish speaking person would call it 😄). With this release I am hoping to make it super easy to start exploring Clojure and ClojureScript. Read more at ClojureVerse https://clojureverse.org/t/the-calva-journey-continues-please-jack-in/4335 Questions and comments and cheering-on welcome in #calva-dev (and/or on ClojureVerse, of course.) ❤️

chris44116:06:42

https://github.com/cnuernber/libpython-clj Some features include: * Python dicts are represented in java as java.util.Map's. stdout,stderr are overridden to write to *out and err, respectively. * transparent access to the python interpreter from any thread. * zero-copy pathways for numpy->tech.datatype tensors and back. So you can intermix heavy processing between clojure and python without pay a large conversion cost. * Decent performance. This is built on the raw cpython shared library so our cost of executing or interoperating with python should not be any higher than anyone else's. * two way bridging. Expose your JVM objects to python and vice versa. I have a simple keras demo where I use Keras from clojure to train a neural network included in the examples. So that proves out keras and tensorflow access from clojure. I think this soundly addresses our ability to load cutting edge math and machine learning libraries.

metasoarous20:06:31

This is amazing @chris441. Thanks so much for your work!

martinklepsch20:06:21

please use threads when replying to postings in #announcements 🙂

metasoarous18:06:37

Sure! Didn't realize that was a thing, but makes sense for keeping the noise down. Thanks for kindly pointing it out to me.

martinklepsch19:06:28

No worries! 🙂