[logistark]

I have the feeling that after Nubank bought Cognitech, Clojure has gone to stagnation. I mean, i feel that promoting and improving Clojure is no longer a priority. And another think that stinks me everytime Cognitech talks about Clojure they have to bring to the table Datomic. They tried to push Datomic on my company long time ago when they do some consultancy job at my company. You can skip all the talk, because is all about how good Clojure is and that is. No any eta about future features, improvements and fix problems of current Clojure users.

Lot of open source contributors have leave the community, because there is any plan on Clojure. I guess that Rick Hickey is happy with Clojure as it is now, and this is the way is going to stay. And still no Java 8 support for CompletableFuture, lambdas interface, Stream api, java.time, no pattern matching, java records, this group feels like a group of old programmers stuck at Java 6 that cannot move forward.

[zcam]

All the things you are mentioning would be nice to haves but frankly nothing prevents us to use these today already, either directly or via a bit of wrapping.

Communication "style" and slow pace is frustrating sometimes but that's a small price to pay for all the positive facets of the language/community. I used to be more critical of these but I don't care anymore, the community is wonderful, the language is very usable and thriving, that's what matters ultimately.

About Datomic, I still don't get why there's no push to open-source the on-prem version, especially since nubank acquisition. I dug onto the internals a few times, contributed to some of the alternatives, read/viewed pretty much everything about it and used it for fun in toy projects, and that thing is just so versatile it makes me angry it's not more accessible and as a result not more popular. It has an incredible untapped potential. Every conj I am holding my breath hoping for a "one more thing" announcement where they'd do just that. The "alternatives" do things either quite differently on too many aspects or lack traction.

[KingMob]

Which of the datalog alternatives do you think is closest? And how much of an improvement do you think a datalog db is over boring Postgres or SQLite? I’ve been weighing what db to adopt.

[zcam]

datalog is just the query language. For me there's no such thing as a datalog db, there are various db that uses some derivative of datalog for querying but that's it. To give you an idea, datomic has other ways to query your data than with datalog alone.

The "closest" to datomic is datahike right now. Crux/datahike/datascript/asami all use datalog in some way or another but they cover different use cases.

[refset]

Just noting that Crux has been rebranded to https://xtdb.com/ (as per https://xtdb.com/blog/crux-to-xtdb-rename/)

[Zak]

It seems to me that these things are a big deal if you're doing a lot of JVM interop and not so much if you're primarily using Clojure-native libraries.

Clojure has futures, lambdas, streams, a community-maintained java.time library, core.match, and Clojure records. I'm sure there would be benefits to adopting the newer JVM-native equivalents, but I'm not as sure they're worth the costs.

[kamaal]

I thought you might be trolling. But then when I looked at the Clojure repo on Github https://github.com/clojure/clojure the last commit was 2 months back. There is some merit in your arguments.

[seancorfield]

Clojure evolves slowly and deliberately -- it gets about one release a year (with various prereleases) so "2 months" between commits isn't a big deal. I consider it a "feature" because it means Clojure is extremely stable -- which is great for us at work given that we've used it in production for a dozen years at this point and have been able to run alpha builds in production safely during nearly all of that time.