iiinova — where your digital self keeps evolving

I’m not that unexperienced person just coming out of school anymore. I know what I want and I’ve helped shape the philosophy of Preact. Seeing the evolution of Preact has been an honor, when I joined we had 300k weekly downloads and now we have over 30 million weekly downloads.

160

As of a certain minor in Preact X and for almost everything in Preact 11 I was the guiding maintainer and it’s been such a surprise realising how I’ve grown over the past years.

(Test) - testutils package with act for hooks by JoviDeCroock · Pull Request #1371 · preactjs/preact

I started eagerly converting our codebase but bumped into act missing from the Preact X alpha at the time, that was my first PR github.com/preactjs/pre... - I’ve done several after that and ultimately became a maintainer. I learned so much working with the wider team.

github.com

Please don't mind where it's located now, I've been trying to make it work. Tests are still a success but the updates aren't synchronous yet, I feel like I am missing something extr...

150

I’ve maintained Preact for a bit over 7 years, I was just getting started in tech and we had this application with horrible performance. I looked around to see what we could do and switching React to Preact was an option. The extensible nature would allow me to tweak rendering.

2191

Can’t wait to see!

030

The wait is finally over: Preact 11 is here!

Thank you everyone for the feedback, pull requests and issues we've gotten over the years, it has enabled this release! preactjs.com/blog/preact-...

preactjs.com

Preact 11 – Preact

061

Hydration 2.0: hydrating while lazy chunks are still loading makes 0 DOM mutations, and lazy components can return a Fragment or null. Returning 0 or more than 1 DOM node in Preact X would introduce hydration mismatches for resumed hydration before all of this.

190

Child reconciliation goes through Element.moveBefore() where supported. Reverse a list of 10 iframes: Preact 10 reloads 9, Preact 11 reloads 0, and the focused input stays focused.

Diagram comparing keyed list reordering behavior in Preact 10.29.8 and Preact 11.0.0. In Preact 11, keyed DOM moves use Element.moveBefore() where supported, preserving iframe state and input focus during reordering. Moving one row, reversing the list, and shuffling the list all reload 0 of 10 iframes and keep focus in Preact 11. In Preact 10, the same operations reload 1, 9, and 6 of 10 iframes respectively, and focus is lost in each case.

2131

preact/compat is now tree-shakeable per feature. A library that only imports memo or forwardRef pulls in ~1.3–1.4 kB of compat instead of ~2.2 kB (−35–40% for single APIs).

1100

Core is 4.5 kB brotli, +75 B over 10.29, while adding Hydration 2.0, minimal-move keyed diffing, moveBefore and createPortal.

Four comparison bars showing Preact 10 vs 11 in terms of published bundle size.

preact +75b at 4459b
hooks +41b at 1419b
compat -177b at 3556b

1100

The npm package went from 1.56 MB to 609 kB (−61%). v11 ships ES modules only, so the CommonJS, UMD and duplicate builds, and most source maps, are gone. The runtime you ship is the same ~40 kB of ESM.

2 bars showing the total package size of Preact, in 10.29.8 it's at 1.56MB and at 11.0.0 609KB which is 61% reduction in size.

1212

Jovi 🐨 @jovidecroock.com · 29/09/2026

Preact 11 is out 🎉 We’ve been hard at work at making our diffing use modern features like moveBefore , leveraging ESM for tree-shaking Preact Compat and ensuring you have a great experience with resumed hydration.

59220

You know what, I'll take a lot less as well

130

Reposted by Jovi 🐨

nate moore @natemoo.re · 29/09/2026

we have no choice but to give the preact team billions of dollars 😌

3486

Reposted by Jovi 🐨

jviide.iki.fi @jviide.iki.fi · 29/09/2026

I didn't want to be the one to say it, but someone has to. Preact 11 will kill all of us.

2545

Reposted by Jovi 🐨

Preact X, a story of stability – Preact

Putting up the PR to move Preact 11 from release candidate to stable feels like something that was a long time coming for me. I've worked on Preact X for the last 7 years, during which we had a different Preact 11 which was based on long-lived vnodes preactjs.com/blog/preact-x

preactjs.com

2224

No codegen needed, you can import the type into your client and use it. See the example

000

It's also quite something to see these AI firms hit record revenue with a lot of the knowledge OSS has put in the world (issues, PR's, source code, ...)

030

Drydock Package Review: read the artifact before you publish

Thank you! I can relate to that, not sure how much of that I'll still work on. If work needs something I can look at it, I've been just building stuff I like like drydock.org but it's all feeling a bit empty lately

drydock.org

Your pull request got reviewed; the published tarball did not. Drydock diffs the exact npm, PyPI, or VS Code artifact and pins supply-chain findings to changed lines.

OSS has been feeling a bit empty for the past year so I am going to be taking it easy for a bit, especially now that my days are really full with a newborn, drydock and work

1120

Releasing Preact 11 with better hydration, moveBefore support and some reshuffling may sound like a boring release but to me it feels like completing something I've thought about for a while, I want to thank everyone for supporting Preact and enabling my journey

1131

I've often referred to Preact X as finished software especially in these last minors... We've published 29 (!!) of them and admittedly, some of it has been a very lonely journey for me

Preact X, a story of stability – Preact

preactjs.com

2224

The newest cards in the directory, most recent first: 154 on file, the latest added 26 September 2026. Follow along with the Atom feed.

Makes me happy that folks are hyped about my Made in Belgium website. Been receiving submissions and it has expanded a lot, go take a look! madeinbe.dev/en/recent

madeinbe.dev

Recently added

081

Jovi 🐨 @jovidecroock.com · 25/09/2026

The way we are currently going, 100% - it's a weird time

010

Jovi 🐨 @jovidecroock.com · 24/09/2026

It's a very hard balance to strike currently given that the role is transforming into a very product-y one when it's viewed by folks in more managerial positions

111

Reposted by Jovi 🐨

It really pains me to lookup GraphQL in search, the technology has gotten a really negative co-notation generally. It was sold the wrong way "solving over-/underfetching" was never the selling point... A product-centric API for all your clients was...

2121

persisted operations solve a lot of the DDos concern as caching and rate limiting becomes easier

011

There's also a lot of misinformation, you don't have to lose out on caching. It's however all under developed, if tinkerers would come back to the tech and make it easy to use persisted-operations, ... it would be far less of a problem I think

030

This also isn't for everyone nor is it for every use case but it really has good use cases and it enables discoverable API's, great generic caches, ...

2121

I was thinking of adding some kind of dependency intelligence to drydock where closed repositories can also log on and see what they depend on, which repositories publish in what way, ... Could also be a good way for OSS maintainers to become aware what orgs use their code

010

I refrain so often from posting to reddit/hackernews due to the amount of negativity it has most of the time. Might be one of the things making the marketing game weaker for mcp-tada, drydock, pracht and preact

050

Jovi 🐨 @jovidecroock.com · 19/09/2026

Agents are really good at diagnosing problems, however next time you are inclined to type "fix it" - read it, look at the source and come up with a fix yourself. Collaborate with your agent, don't trust it for opinions, 9 times out of 10 the solution it would bring is bloated

161

It should yes, all though I wonder whether this increase in min release age will slow down compromise detection. Most detection is reactive, drydock was my attempt at making it proactive but… npm does not give us much to work with.

110

Jovi 🐨 @jovidecroock.com · 19/09/2026

One thing I find dangerous in the current MCP ecosystem is the tendency to advertise npx -y <mcp> which will always get the latest version, compromised dependencies,... included

220

Jovi 🐨 @jovidecroock.com · 19/09/2026

Drydock now has release memory, it will demote unchanged context risk when the package has been reviewed before. I noticed that a lot of deterministic findings would keep surfacing, when they've been checked already, they should not resurface! github.com/JoviDeCroock...

github.com

GitHub - JoviDeCroock/drydock: Review the artifact before you publish it: diffs the built npm/PyPI package against the last published version and pins risk signals to the changed lines

Review the artifact before you publish it: diffs the built npm/PyPI package against the last published version and pins risk signals to the changed lines - JoviDeCroock/drydock

000

One final tip for folks using OIDC publishing, turn on "Use immutable subject claim" in your OIDC GH repo settings

000

Finally, cache poisoning issues can be prevented by not leveraging caches inside of your publishing workflow. I hope this can help some people out, I have also added this as a skill at github.com/JoviDeCroock...

github.com

110

A compromised action can be prevented by always pinning the action by SHA or ensuring the repository uses immutable releases and pinning to a release. Tip: enable "Require actions to be pinned to a full-length commit SHA" on your repositories

100

Drydock Package Review: pre-publish package security

A malicious actor pushes code to the repository which can result in a direct publish or some hidden code that gets later published by the author. With stage publish we can prevent this, we can leverage a tool like drydock.org to verify before it hits the registry

drydock.org

Review the exact npm, PyPI, or VS Code artifact before publication. Drydock diffs built package bytes and pins supply-chain risks to changed lines.

100

Direct publishes by a malicious actor can get verified, with current provenance we can prove that a publish comes from a verified source (repository, ci-run, ...) We sadly can't tell npm to disallow direct publishing and only allow i.e. stage publish with multiple approvals

100