Clojurians Log v2

LevelDB is ready to rumble: https://github.com/alekcz/konserve-leveldb

did you do it from scratch?

Not entirely. I reused a lot of the code from the other konserve backends I worked on.

it did take a while get the linear storage layout sorted though

And RocksDB https://github.com/alekcz/konserve-rocksdb

really cool. I want to get into backends soonish as well. I want to write a cassandra-konserve sometime...

Happy to chat about how I approach it when you're ready to get started

will do, thanks!

You mean Reeeeady to REPL! 🙂

I am trying to use datahike server, and I can get to the main to run, but I can't get to create a single transaction

I basically stop at the coercion all the time.

Would you have an example ?

I get a java.class.Assertion error.

Moreover, I think the uberjar from the repository does not hold the latest dependencies and breaks.

So I made it work, but only with application/edn. I wondered if it would be possible to make it work with json?

Hi @neo2551 thanks a lot for testing it!

In case it is not mentioned on the repo it is still in early stages.

So it is strange: the datahike server works as expected if I use application/edn, but once I communicate application/json, I have a bad entity

[{:age 35, :name David} {:age 20, :name Myriam}] 20-09-23 19:34:40 amd-3700x ERROR [datahike.db:1054] - Bad entity value 35 at [nil nil :age 35 nil] , value does not match schema definition. Must be conform to: (= (class %) java.lang.Long) {:error :transact/schema, :value 35, :attribute :age, :schema {:db/unique :db.unique/identity, :db/valueType :db.type/long, :db/cardinality :db.cardinality/one, :db/ident :age}}

We will be working on JSON support very soon

No problem, I feel honored to help you 🙂

and I think there is some kind of problem with JSON right now. I did not investigate it further but some kind of types or namespaces will not be translated correctly right now.

I also had a problem with a jdbc backend: I had to add the dependency manually in datahike-server.core, otherwise it would no be found.

But I will continue to study datahike, but it seems my team are going to try to use it in a polyglot setting as well (python and R).

Thanks a lot for your work.

Great David. That is very good to hear. We are working a lot on improving Datahike and Datahike server is a very important project for us.

If you can describe your problem with JDBC a bit more in detail I could investigate it some more. But we will do some more testing with JDBC as well of course.

I tried to use Datahike server with a jdbc backend [if you could migrate to deps.edn it would be nice as well], and the Uber jar did not want to start because Datahike-jdbc.core was not required (as it was AOT compiled I guess?).

I solved it by requiring Datahike-jdbc in main.

Actually we do migrate to clj-tools with deps.edn for the CLJS-support that is in the works.

🙂