Holochain

Sam and Anita, the fantastic team at hAppenings, have published a list of updates from around the Holochain ecosystem. Some light springtime reading. Catch up on the latest Holochain community news, events, and releases in the hAppenings newsletter. Grab a coffee and read it here: happeningscommunity.substack.com/p/spring-break… #Holochain #Web3Community #OpenSource

Iroh Networking on Holochain: How it started: 🤔 How it's going: 🚀 0.7 got a new network layer powered by Iroh, promising zero-hassle bootstrapping and graceful recovery. It worked so well we ported it back to 0.6. Now it's powering the Holo & Unyt technical migration tests and feedback is glowing. #Networking #DistributedSystems #Iroh

P2P and blockchain — why not both? 🦆🤝🐟 Holochain is anti-blockchain in the same way a duck is anti-fish. (In other words, it's not). Holo and Unyt are currently testing ETH bridging: Lock HOT, get a HoloFuel credit. It has a lot of moving parts, but it’s working. Move smoothly between decentralised ecosystems without the drama. #Web3 #Ethereum #Holochain #Interoperability

Tried the Holochain 'Getting Started' dev experience lately? We broke it. 💥 But we fixed it! 🛠️ 0.6 is currently in an awkward teenager stage. Scaffolding an app gives you 0.6.1—not perfectly stable yet, but you get the massive Iroh transport upgrades and a pile of bug fixes. Now with 0% broken stuff 🤞 developer.holochain.org/get-started/ #DevExperience #Holochain #Rust

Okay, it's time to vibe-code that hApp you've been dreaming about. 🤖✨ Try @soushi888's new skill for Claude Code. Go from an "empty folder" to "decentralised awesome for my friends and me" in record time. Try it, break it, and leave detailed feedback on what it still needs: github.com/Soushi888/holo… #AI #ClaudeCode #Holochain #Web3Dev

The Holochain 0.6.1 release projection just dropped from five weeks to TWO weeks. 📉⏩ As we’ve mentioned before, completion percentages and projections are a snapshot, and a lot can change in a single week. Keep riding the roller coaster with us! 🎢 #Holochain #RustLang #DevUpdate #OpenSource

Sprint velocity has been going up steadily — we're up to 34 points per week on average now. Why? It may be due to a new team member joining, or productivity boosts from judicious use of AI developer tools. We also have a hunch that it has something to do with our switch from rolling kanban to disciplined sprint project management — reduced cognitive overhead for devs, more accurate time projections (and as a side benefit, more clarity and confidence for projects building with Holochain!)

Community member Sacha Pignon (@soushi888) has released a 1.0 beta hApp development skill for Claude Code, Cursor, GitHub Copilot, Augment, and Codex. Its skills take you through the full hApp lifecycle, from empty folder to distributable binary. github.com/Soushi888/holo… A v2.0 is already planned, with support for ecosystem libs like hREA, AD4M, and holochain-open-dev.

Q: Why does 0.6.1 look like it's getting *less* complete over time? @H_O_L_O_ and @unytco are actively alpha testing right now. Bug reports come in multiple times a day. When that happens, the scope for 0.6.1 grows, and the completed percentage drops. That's a good thing! It means we're finding bugs before they hit production.

Why does Holochain 0.6.1's release date keep retreating? The biggest reason is that these are teamwork weeks, not calendar weeks, and our priorities shift per sprint. This week might be more Wind Tunnel; the next week might be more Holochain. Next, critical bugs discovered by @H_O_L_O_ , @unytco , and other projects expand the scope. Projections are also based on a 13-week rolling velocity average, so next week might be faster or slower than last. Finally, each sprint sees re-shuffling of tasks, especially as a release get closer. Holochain 0.6.1 may be out sooner than you think. tl;dr: our roadmap is a snapshot of our dev team's actual work. Hold the numbers lightly. blog.holochain.org/introducing-th…

In January our dev team switched from kanban project management to traditional two-week sprints. The goal was to increase clarity and reduce surprises. Is it working? Yes! Armed with a wealth of velocity data and disciplined scopes for each sprint, we now have more confidence about completion times, which gives projects like @H_O_L_O and @unytco more confidence in us.

Right now the roadmap is reporting 34 points per week. Does that mean we're faster than ever? Yes and no. The velocity number is steadily increasing, and that's great. (We think it's due to a new team member and project management efficiency boosts.) But to interpret it right, look at it as a rolling average of the last 13 weeks of real work. Efficiency gains push it up, holiday seasons push it down. The averaging algo smooths these blips and doesn't perfectly predict what next week is going to look like. Its job is to predict a typical work week. Roadmap: holochain.org/roadmap

When you look at our roadmap, do the 'points' confuse you? Do you wonder what they mean? Software time estimation is notoriously hard. Great teams get it wrong. The best teams don't do it at all. Instead, they (we included) use evidence-based scheduling. Estimate the complexity of each task using points, gather data over time, then use past performance to predict how many points you can get done in the next two weeks. It's imperfect, but it's empirical, and it just works better. Roadmap: holochain.org/roadmap Evidence-based scheduling: joelonsoftware.com/2007/10/26/evi…

Thanks to everyone's support, we're able to run large-scale tests. We've got over 1700 peers in this run; other runs have over 2700! holochain.github.io/wind-tunnel/ru…

1700+ peers. 2700+ peers. Longer runs now, watching what degrades over 30 minutes instead of 10 seconds. We're not measuring "performance is great." We're measuring where it breaks and how it breaks. We want the bad news. That's what you learn from.

Sweettest Github: github.com/holochain/holo…

Sweettest, our internal testing framework made ready for you to use, replacing Tryorama in our dev tools. Scaffolding now generates Sweettest test cases. Same language as your zomes (Rust), so you're not context-switching into TypeScript. You write the test in the language you're already thinking in. No official docs yet, but the source is well-commented. (Link below)

Getting Started Guide is working again after some dependency mismatches. Changes: Playground is deprecated/removed. Tryorama has been replaced by Sweettest. 0.6.1 is still in RC mode, and we're not promising it's stable yet. But if you're trying to learn the basics, you can actually try them now instead of hitting dead ends.

What are we building in public right now? Machine-to-machine payment infrastructure that autonomous systems can depend on. Multi-currency transactions without exchange layers. Compliance-grade payment automation.

Testing feedback is surfacing fixes we didn't anticipate. Some fixes are rolling into 0.6.1. Some are still being figured out. This is how it works when the system is under real load. The scope grew because the testing found work worth doing — work that's necessary for Unyt to be a production-quality system.