Clojurians Log v2

prod link to the https://demo.clojure.net/java.lang.management.ManagementFactory$getPlatformManagementInterfaces/~map,(:hyperfiddle.navigator6$selection,~vec,(ManagementFactory$getPlatformMXBean,(java.lang.Class$forName,'java.lang.management.RuntimeMXBean')))/~map,(:hyperfiddle.navigator6$selection,~vec,(runtime-classpath-entries,(java.lang.management.ManagementFactory$getRuntimeMXBean)))/~map,(:hyperfiddle.navigator6$selection,~vec,(:path,(dustingetz.nav-jar$jar-by-path,'fiddle.jar')))

Can I see a screenshot of your Rama Pstate explorer

I couldn't figure out how to inspect Rama modules to see Pstate schemas and make a generic explorer (not sure if that's possible or even makes sense given Rama's flexibility for data modeling). It isn't much, but here you go! Also note that right now all my data is under a specific known top-level key in Rama. For a more serious pstate, querying might be more difficult since you typically need to know the first key in the path query.

where is? 🙂

https://www.clojure.net/ and https://www.youtube.com/watch?v=aLIMLbAWSfY

I found https://github.com/hyperfiddle/hyperfiddle-starter-app/blob/main/src/dustingetz/main.clj helpful for reference when running locally. I felt weird about "connecting" to the http://clojure.net cloud proxy without discussing with Dustin first, but guess that is the point that it's just so easy to expose data? It's so easy that we'll need to be careful about exposing data from 3rd party APIs and not violating license terms 😅

if the tool proves useful to the community we will offer private subdomains and such, make it competitive with any old low code tool like Notion/Retool/Google Sheets, same security model as that

does that resolve your concern?

Yes, it isn't really a concern but just something I started to think about as soon as I had an explorer agent working with my data. I am probably more interested in a self-hosted model anyway, or at least I enjoy having that option to support internal environments for testing etc. I'm still wrapping my head around what this could unlock and where the vision is going and it's fun to be along for the ride

I am also curious if there are plans to support more reactive rendering models, like streaming updates from data sources that support that via electric (rama async pstate proxy)? Or is that too electric-specific and "in the weeds" for the scope of hfql/clojure.net?

There is a self-host capability but it adds friction (hosting and docker and url coordination and updates and security and frontend builds etc), our hypothesis is that most of the value of this tool is in eliminating all that friction. Businesses need to self-host but they have platform teams dedicated to operating infrastructure, security updates, kubernetes, etc

this repo is configured for self-host via loopback if you want to try it https://github.com/hyperfiddle/datomic-browser but again it damages the value prop

Yes, I am content to handle getting it running and hosting. I've been doing that so far for electric with no issues. I don't think the build stuff is that hard to figure out if you've been doing full-stack clojure for a while and understand the constraints around jvms and classpaths

Reactive/streaming data sources – absolutely on the roadmap, we just aren't ready yet, and furthermore most businesses have request/response OLTP-shaped database and API workloads. At that point, this tool can be a Discord client, which would be insane

Awesome! Yeah that's what I expected and makes perfect sense. I'm glad it's on the roadmap though 🙂

awesome

claude skills for HFQL are coming soon! and like 50 examples