hyperfiddle

Dustin Getz (Hyperfiddle) 2025-07-09T13:10:03.840079Z

With Electric rapidly maturing, we are now starting to slow-launch Hyperfiddle: • https://github.com/hyperfiddle/hyperfiddlehttps://github.com/hyperfiddle/datomic-browser What do you think?

😮 3
🥳 6
🚀 18
hrtmt brng 2025-07-12T18:50:02.163929Z

My first impression of Hyperfiddle: Big surprise for me! I spent weeks, months, years reading and trying out so much, because so often I need a configurable generic data browser/viewer/editor. And this just does not exist. After years, I finally gave up. The closest to my needs, what I found were Jetbrains MPS, the XML tools (like XMLMind) and Limbas. Hyperfiddle seams to be what I am searching for a decade already. You said, Electric is from experts for experts. I hope, Hyperfiddle will also be for non-experts.

🙂 1
hrtmt brng 2025-07-12T19:21:16.375159Z

But one side note. Success will not come automatically. Someone would believe, if you have a perfect product for a need, which everybody has, you eventually will be successful. That is not the case. As an evidence (if you are interested), you should look at Jetbrains MPS. That is an awesome tool for Domain Specific Languages. With very limited skills you can build impressive editors for your DSL. The author even believes, this is the killer app for DSL Editing. And he is not an unknown person. You would think, this must be a success, because so many people use something like DSLs every day (config files, data editors, content editors). The tool is is even Open Source. But nobody is interested in that tool. I showed it to a few colleagues. Not a single person is interested. People are paid to do their job. If they use bad tools they are still happy. I think you will often run into a psychological phenomenon: You show something to people, which actually solves their problem. But the audience does not even realise that what you show them has something to do with them.

telekid 2025-07-10T13:18:08.414899Z

Exciting to see, a long time coming! Feedback from a distance: 1. I think a 60s video would be a better format than screenshots at this point, given the level of interactivity that you're providing. I wouldn't worry about production value at all, just clearly conveying the idea. 2. What would a postgres browser look like? Everyone knows what postgres is, so it would be an easier conceptual leap for people. 3. After reading it a few times, I think the big idea here centers around very rapid development of high-quality back office / operations applications. If that's the angle (and I think it's a good one), I'd hit it harder. 4. In general (but not knowing your target audience), I think this is too technical / in the weeds. The key business value is "build this UI with one developer and 300 lines of code" or similar.

telekid 2025-07-10T13:20:00.979099Z

(That's how I would have pitched it when I was at Flexport, a company that had ~3m of operations-facing react code back in 2019.)

Dustin Getz (Hyperfiddle) 2025-07-10T13:20:16.700179Z

postgres browser is coming, we have to figure out how to host it

🔥 3
Dustin Getz (Hyperfiddle) 2025-07-10T13:20:41.226399Z

its just more work than datomic, and datomic is a specific market that we happen to like a lot for internal reasons

telekid 2025-07-10T13:22:24.166809Z

Makes perfect sense. Cool to see hyperfiddle-the-product emerging in the wild, rooting for you!

👍 1
Jim Duey 2025-07-18T13:11:42.547659Z

Just now catching up to this. Exciting! Cheering you guys on.

🙂 1
Andrew Wilcox 2025-07-10T02:13:07.105759Z

H1's are the hardest to write... I don't know what "high fidelity" means. The pain point you're addressing, as I understand it, is that writing enterprise frontend code is a pain because A) it involves writing a lot of tedious REST APIs, and B) the result is lousy (slow and brittle). "Easily write high-performance enterprise frontends"? Negative emotional reaction to "Stop treating your ops teams badly". No one likes to be told that we're treating people badly. (In our own internal story about ourselves, we're doing the best we can.) Maybe something like "Make your ops team more efficient, freeing them to focus on core business projects."

👀 1
➕ 1
Andrew Wilcox 2025-07-10T02:26:05.495049Z

"Hyperfiddle does NOT depend on Datomic". I imagine the person paying for this is going to be the engineering manager or the CTO. It might be an individual developer who brings it to their manager's attention ("hey, this is cool, maybe we should pay for this"), but it's the manager (I imagine) who is going to be signing the invoice (your paying customer). The manager may not know what Datomic is, and might not know why someone might think that Hyperfiddle would depend on Datomic. I suggest expanding on your headline: "Here is an example of connecting to the Datomic object database." OK, if I were a manager and I told one of my engineers to write an object database browser and they didn't have Hyperfiddle, what problems would they run into? In this particular example of an object database browser, what does Hyperfiddle solve for my engineer?