jackdaw 2019-07-30 | Slack Archive

@quest examples/dev/user.clj

👍 4

D'oh! I should've grep'd the whole repo. Thanks!

Happy to help

1st time ever, just got a message from Clojure server -> Kafka topic -> Onyx job. Awesome! Next up: writes to Datomic. (Just excited and wanted to share -- the help is appreciated!)

xi13:07:32

🎉

xi13:07:56

Out of interest, why are you using Onyx?

Quest17:07:51

#Also sent to the channel

Working on a pure Clojure+CLJS ecosystem that has a streaming data-model from the ground up -- a pure event sourcing architecture to support reactive subscriptions from any clients over any data. The computational model that Onyx provides looks more powerful than building on top of Kafka streams. Namely: • lower latency (KStreams has ~100ms builtin?) • batch processing • better native Clojure experience • in memory CLJS implementation https://github.com/onyx-platform/onyx-local-rt My goal is to achieve dynamic distributed datatypes + reuse of the data-schema & relevant business computations across all connected runtimes. In addition to the business goals, I want to put a distributed Clojure REPL into the hands of every user. I'm open to reconsideration of KStreams -- my goal is to maximize the power of the distributed runtime at any cost, including my own time. This project is nearing year 3, and I recently left my nice Clojure job to pursue it fulltime.

👍 4

cddr17:07:59

I definitely dig onyx. Investigated it a few years ago, (before kafka streams was even released) and would probably be using it if these types of decisions were up to me. I think the main concern at that time was that it was a bit of a one-person project (albeit a fantastically talented person).

Quest17:07:51

Your concern ended up being valid -- the creator admits that the project is currently in maintenance mode. Easier decision for me since I'm not betting any existing business on it 😅 -- & if I succeed then I'll fund Onyx dev

cddr15:07:50

Hey @slack1274, I'll deploy a release today. Seems like there's a few things pending. I had been trying to fix https://github.com/FundingCircle/jackdaw/issues/146 before making a release but it is proving to be quite stubborn and I'm struggling to get a good run at it. Also remember that you can always publish a snapshot of your fork under your own clojars group by updating the project.clj and running lein deploy clojars.

xi16:07:04

Thanks! For the merges as well. I actually have published my own, org.clojars.lxsli/jackdaw:0.6.6a 🙂

xi16:07:56

It's really nice to have an official release though

cddr16:07:16

Release - The magnificent 0.6.7 Added • Allow specification of :deserialization-properties (#157) • Back-fill a few tests of jackdaw.client.partitioning (#165) Changed • Upgrade Clojure version to 1.10.1 (#159) • Partitioner in test-machine write command updated to match streams (#139) • Reformatted all the code using cljfmt (#173) Fixed • Delete duplicated tests (#165) • Documentation/Examples fixes (#166, #168) • Do not assume result of executing command is a map (#164) • Supply key-serde as well as value-serde in aggregate methods (#172) 👏👏@AndreaCrotti, @MicahElliott, @AlexVPopov, @yogidevbear, @funderbar, @lamp, @alee 👏👏

16

dharrigan17:07:34

you should announce in #announcements too 🙂

👍 4

Quest17:07:51

replied to a thread:@quest `examples/dev/user.clj`

Working on a pure Clojure+CLJS ecosystem that has a streaming data-model from the ground up -- a pure event sourcing architecture to support reactive subscriptions from any clients over any data. The computational model that Onyx provides looks more powerful than building on top of Kafka streams. Namely: • lower latency (KStreams has ~100ms builtin?) • batch processing • better native Clojure experience • in memory CLJS implementation https://github.com/onyx-platform/onyx-local-rt My goal is to achieve dynamic distributed datatypes + reuse of the data-schema & relevant business computations across all connected runtimes. In addition to the business goals, I want to put a distributed Clojure REPL into the hands of every user. I'm open to reconsideration of KStreams -- my goal is to maximize the power of the distributed runtime at any cost, including my own time. This project is nearing year 3, and I recently left my nice Clojure job to pursue it fulltime.

👍 4

2019-07-30

Channels