洪 民憙 (Hong Minhee) :nonbinary:

@evan@cosocial.ca Thanks for putting this together, Evan. While implementing FEP-6757 in Fedify, I found a few errata: schema:license and cc:license use IRIs that don't match Schema.org and ccREL, a couple of example typos, and an open question about @type: @id coercion. I filed them on your tracker: https://codeberg.org/evanp/fep/issues/52

Fedify 2.4.0 is out. The main addition is support for FEP-ef61 portable objects. A Fedify server can now host actors whose identity is a DID instead of a domain, serve their signed objects through /.well-known/apgateway/, and verify the portable objects of other servers. I tested it against tootik and Mitra, and wrote up where Fedify's profile differs from the FEP and what it leaves out. The release also adds application-defined background tasks, the ActivityPub Media Upload extension, per-request inbox reports, 410 Gone for deleted objects, and four new packages (AdonisJS, Netlify, PGlite, and interaction controls). Some defaults changed, so please read the upgrading section before you update, especially if you federate with software that uses FEP-ef61 compatible identifiers. Release notes: https://hackers.pub/@fedify/2026/fedify-2-4 Thanks to everyone who contributed to this cycle, and to @silverpill@mitra.social for patient answers about the FEP. #Fedify #ActivityPub #fedidev #FEP_ef61

One thing I keep coming back to when comparing #ActivityPub and AT Protocol: ActivityPub is basically a convention on top of the #web stuff we already have. An actor is a JSON-LD document you GET. An inbox is an endpoint you POST to. If you already run a website, you can bolt this on without changing the site's basic shape. Ghost didn't set out to join the #fediverse, and then years later it could, just by adding an endpoint. AT Protocol feels less like that to me. Running a PDS means a signed repo, a Merkle search tree, a firehose, a DID. It's not a layer you add to an existing site. It's closer to standing up another backend beside it. I think that's why ActivityPub keeps making sense to me. You can join later. You don't have to have built the whole thing with federation in mind from day one. #ATProto #ATProtocol #fedidev

A Chinese developer, @21018365486@o3o.ca, wrote about adding ActivityPub to her blog using Fedify. Thank you for using Fedify! https://blog.yunyi.beiyan.us/posts/migrateToFedify/

Okay, now it's merged! https://w3id.org/fep/22cd

And that's how I ended up with about 30 domains. 😂

RE: @dysfun@social.treehouse.systems

A friend of mine just launched Kosmo, a new #fediverse service out of Korea, built around the dōjin (同人) scene: fan works, fanzines, that whole self-published creative culture. It's live now at kos.moe, in open beta and Korean-language only for now. It speaks #ActivityPub, so you get the basics: posting, following, boosting. It also has emoji reactions, and one account can hold several public profiles. They say profile tags are coming next, meant to express something about a profile's identity and usable for search and muting too. It's still early, and they'd genuinely appreciate feedback from anyone who gives it a try. #fedidev #doujin #dōjin #subculture #otaku RE: https://planet.moe/@kosmo/117014381597184627

Quick update: implementation of FEP-22cd has started in Fedify (PR #1038). Once it's reviewed and merged, it'll ship in Fedify 2.4.0, which I'm aiming to release within September.

#Mastodon's #ActivityPub inbox treats object types in tiers. Note and Question are supported directly. A second tier, Article, Page, Image, Audio, Video, and Event, gets converted to a title, a summary, and a link; content itself doesn't show. Anything outside both tiers is dropped on delivery, which is a large part of why so much of the fediverse federates as Note regardless of what the object actually is. mastodon/mastodon#24079 asks for content to show up whenever it's present and the mediaType is supported, independent of type. Open since 2023, and as far as the thread shows, no core team engagement beyond one offhand remark that they're not happy with the status quo either. Reading the issue text closely, it's a bit ambiguous whether it covers the drop tier or only the convert-to-title-and-link tier. I left a comment asking for clarification, since the fix looks very different depending on the answer. If you build on ActivityPub, a comment or a reaction there is worth more than another workaround. #fedidev #fediverse

The more I work on Hollo, the more its Mastodon-compatible API feels like a straitjacket. I chose it for a practical reason: people could use existing clients, and I wouldn't have to build a frontend too. I still think that was the right call at the time. The problem is that clients can only expose what Mastodon's API represents. I keep coming up with ideas, then asking whether the API has anywhere to put them. If it doesn't, I have no way to get the feature in front of users. ActivityPub API (so-called “C2S”) might be the eventual answer, but it doesn't help much today. I don't know of any clients Hollo users could realistically use with it, and adoption is still limited. So I keep coming back to the idea that Hollo will need a frontend of its own. I'd keep the Mastodon API for existing apps. I'm not especially eager to take on a whole frontend, but I don't see another way to try ideas that don't fit Mastodon's model.

I write my name in Chinese characters: 洪民憙. In English I still include them, writing Hong Minhee (洪民憙) rather than the romanization alone.

Part of it is meaning: Chinese characters carry it, hangul doesn't. Mostly, though, it's about the direction of loss. You can derive a Korean or romanized reading from 洪民憙; you can't go the other way. Given 홍민희 or Hong Minhee alone, there's no recovering which characters were intended.

I don't much mind how my name gets pronounced. If a Chinese speaker reads it as Hóng Mǐnxī, or a Japanese speaker arrives at something different, that's fine. The characters are the same. Classical Chinese once worked this way across East Asia: one written text, many readings.

What I care about is the notation.

I once gave up on building a federated microblogging app because the ActivityPub work swallowed it. That frustration turned into Fedify, and the app I shelved eventually shipped as Hollo, built on top of it.

I wrote up where implementing ActivityPub actually hurts, and why it doesn't have to.

RE: @fedify@hackers.pub

光栄なことに、DrFed（@drfed@hackers.pub）チームを代表して、APC（@APC@mastodon.social、Association for Progressive Communications）の『自由な未来のインターネット作り』（Building a Free Internet of the Future）インタビューシリーズにお答えすることになりました。（英語） https://www.apc.org/en/news/drfed-reducing-complexity-and-enabling-more-people-shape-fediverse

中國語(중국어)로 「키보드 워리어」를 「鍵盤俠(건반협)」이라고 부른다는 걸 오늘 알게 되었다. 너무 재밌는 말이라 韓國語(한국어)로 輸入(수입)하고 싶을 程度(정도)이다.

I get the impression that people are more open to generative AI like LLMs in North America than in Europe, and even more so in East Asia than in North America, and I'm curious as to why that is.

Just created my CodeFloe account!

https://codefloe.com/hongminhee

RE: @codefloe@social.tchncs.de

SQLAlchemy is hands down the best ORM I've ever used—I actually used Python just because I wanted to use SQLAlchemy. I use TypeScript as my main language now, but I still miss the SQLAlchemy + Alembic combo. Drizzle ORM is decent enough, but still…

I've been using mise a lot lately, so when I'm bored, I've been typing mise upgrade instead of sudo dnf upgrade -y.

Small fedi badge update: the service has moved to a new domain. Please update your badges to use fedi-badge.minhee.org instead of fedi-badge.deno.dev. The old domain will stop working in July. I also shipped a couple of related changes. Badges can now show the logo of the software an instance is running, like Mastodon, Misskey, Pleroma, and a few others, detected via NodeInfo. If the software isn't recognized, it falls back to the generic fediverse logo. The service also moved from Deno Deploy Classic, which is sunsetting, to Cloudflare Workers.

It is an honour to have been interviewed on behalf of the @drfed@hackers.pub team for @APC@mastodon.social's Building a Free Internet of the Future series. https://www.apc.org/en/news/drfed-reducing-complexity-and-enabling-more-people-shape-fediverse

I didn't realize this, but you can use @hollo@hollo.social with the latest version of the @ivory@tapbots.social app! However, notifications aren't working just yet. I'll need to look into it soon and try to make it fully compatible.

Some amusing timing: Vercel Labs just published scriptc, with a strikingly similar surface goal: compile JavaScript/TypeScript ahead of time to a small native executable, with no Node.js or V8 embedded. Apparently someone else wandered into roughly the same problem at roughly the same time. The design underneath is almost the opposite bet, though. scriptc runs the real TypeScript 7 checker and treats its results as facts, lowering them into a typed IR with monomorphic arrays, exact records, tagged unions, and so on, then through LLVM or a C fallback. Anything outside that model, including any and plain JavaScript packages, needs --dynamic, which links QuickJS-ng as a separate island with its own heap and microtask queue. Values are copied and checked at the boundary. Its correctness evidence is 800+ differential tests against Node.js, rather than a spec-conformance claim. Oseo trusts none of that. A value annotated number may still be a string at runtime, and the compiled generic path has to handle it without an escape-hatch VM. Typed and untyped code use the same native generic-plus-guarded model, and the target is versioned ECMA-262/test262 conformance rather than matching Node.js behavior. scriptc trusts the checker and isolates what it doesn't trust in another engine; Oseo trusts nothing, but has no second engine to isolate it into. scriptc is already a much broader, more mature Vercel Labs project. I just find the near-convergence from opposite directions rather pleasing. Filip Jerzy Pizło, Fil-C's author and someone who makes dynamic languages fast for a living, left a fairly harsh critique of scriptc on Hacker News, worth reading in full. His core complaints, QuickJS-ng as the any fallback and floats-only numbers without integer inference, land on exactly the two bets Oseo refuses to make: a guard failure here falls through to already-compiled generic code, not an interpreter, and there's no static gamble on numeric representation to begin with.

I know it's healthy not to take criticism of my work personally, but sometimes the criticism I get about my work still hurts quite a bit.

I watched Star Wars: The Mandalorian and Grogu today, and I completely fell in love with Grogu. He is so cute!

@fedicat@pc.cafe Yeah, the quote post collision was a clear example. Whether it's deliberate or just indifference, the Mastodon team doesn't seem to treat the API as a shared surface that evolves across implementations. Hard to build on top of that.

@smallcircles@social.coop Fair, and I'll be honest, I haven't read the linked post yet, it's a long one and I want to give it a proper read rather than skim it. I'd like to be more involved in the standardization side of this too. Language is a real hurdle for me there, participating in that kind of discussion in a second language is harder than it looks, so bear with me if I'm slower to engage than I'd like. That said, writing this case up as a FEP seems worth trying, and I might take a shot at it.

This is like magic, absolutely awesome. I'm really looking forward to PostgreSQL 19.

RE: @lobsters@mastodon.social

It might be a kind of Stockholm syndrome, but I often get the urge to code in programming languages that tormented me in the past, like PHP or C#.

@_elena@mastodon.social Thanks! Yeah, this kawaii cover would be one of its selling points!

If you're interested in this work you can subscribe umbrella issue #288.

Just filed sub-issues for FEP-ef61: #826 Support ap: and ap+ef61: URI schemes in vocabulary codecs#827 Resolve did:key verification methods for Object Integrity Proofs#828 Canonicalize and compare ap:/ap+ef61: URIs#829 Support cryptographic origins for ap:/ap+ef61: URIs#830 FEP-ef61 vocabulary terms#831 SHA-256 digestMultibase helpers for portable media#832 FEP-ef61 proof policy for portable objects#833 Compatible FEP-ef61 identifiers for portable objects#834 Gateway dereferencing for portable vocabulary references

基本的(기본적)으로는 韓國語圈(한국어권)도 日本語圈(일본어권)과 大同小異(대동소이)하긴 한데… 다만 韓國語圈(한국어권) 聯合宇宙(연합우주)는 全般的(전반적)으로 LGBTQIA+ 親和的(친화적)이라는 印象(인상)은 있다.

RE: @dev@post.naskya.net

@django@social.coop That would be great. A themeable C2S PWA that Hollo could reuse as its own frontend is exactly the kind of thing I've been hoping for. Looking forward to it.

@evan@cosocial.ca Yeah, exactly. I'm thinking Hollo's own frontend could use the ActivityPub API internally.

@fedicat@pc.cafe That's fair. I know other platforms extend the Mastodon API, but I've been reluctant to do the same. It doesn't feel like a surface that's meant to be extended, and I'd rather not pile more non-standard behavior on top of an API that's already a moving target.

Upyo 0.5.0 is out: a cross-runtime email library for #JavaScript and #TypeScript, one Transport API across SMTP, JMAP, Mailgun, SendGrid, SES, Resend, Plunk, and now Lettermint.

This release makes failed sends structured (category, retryable, HTTP status, Retry-After) instead of just a string, adds a retry transport with backoff and jitter, and brings OAuth 2.0 to SMTP for Gmail and Outlook.

https://github.com/dahlia/upyo/discussions/29

@crepels@mastodon.social Hi, it seems like ActivityPub.Academy is down right now. Could you please check on it? Thanks as always for your help!

@dansup@mastodon.social I really wanted to go too, but I won't be able to make it because the dates overlap with COSCUP 2026. 😭

@stefan@stefanbohacek.online Thanks!

@DoomHammerNG@mastodon.social I guess you could buy an e-book copy?

@mistheart@masto.es Thanks!

@JonasJRichter@mastodon.pnpde.social Thanks!

@raisondetredev@mastodon.de Thanks!

@bgl@hackers.pub Different goals, really. AssemblyScript is its own statically-typed language that happens to look like TypeScript: it accepts a strict subset of the syntax, wants explicit types like i32/f64 rather than dynamic values, and doesn't implement JavaScript's actual object and prototype semantics. You write new code specifically within those constraints to get compact WebAssembly out. Oseo is trying to keep full ECMA-262 dynamic semantics, arbitrary shapes, any, all of it, and still reach native code, with types and JSDoc used only as opt-in hints that fall back to a compiled generic path whenever they're wrong. AssemblyScript trades JavaScript semantics for a clean path to WebAssembly; Oseo is betting it doesn't have to make that trade. Similar-looking syntax, close to opposite premises.

@evan@cosocial.ca Thanks!

@evan@cosocial.ca @dansup@mastodon.social If tags.pub has such feature it would be really great! I want to integrate it with Hackers' Pub!