洪 民憙 (Hong Minhee)

@silverpill Fair, but a lot of the celebrated ones (network effects, data gravity, ecosystem lock-in) work by making you hard to leave. The competitive advantage and the trap are often the same mechanism.

The fediverse is full of real life superheros.

Let's shed some light on a few:

@aral - Building https://kitten.small-web.org + @gazaverified

@quillmatiq - Building https://brid.gy + @anewsocial

@silverpill - Managing FEPs + @mitra

@julian - Building @nodebb

@jesseplusplus - Building @frequency

@jaz - Building @teamtoot + @iftas

@thisismissem - Building https://fires.fedimod.org

@benpate - Building https://emissary.dev

@koen - Building https://procolix.eu

Show em' some ❤️

This article is also a great example of confirmation bias. The bias being that AI is destroying the planet, so this article *must* be correct and factual.

The researchers are trying to attribute the heat increase to the servers running within the data centers, when we know from other research that this likely isn't the case. We have other explanations, they just didn't fit the narrative for this story.

Many datacenters have switched to closed loop cooling systems & have realised that the heat the servers produce is actually a resource they can sell, whether through heat exchange programs or municipal heat networks, or to industrial users who often need heat. Not capturing that heat your datacenters are producing would literally be throwing money away, burning cash.

So I'm seeing this article from NewScientist going around here too, and I'm begging people to use their critical thinking skills for just one minute. This article is misinformation, it is correlation without causation, and honestly, New Scientist should be ashamed at publishing it.

https://www.newscientist.com/article/2521256-ai-data-centres-can-warm-surrounding-areas-by-up-to-9-1c/

They've now put it behind their paywall, it's quite literally clickbait to convert people to subscribers. Let's look at some of the claims:

”Marinoni says that areas including the Bajío region in Mexico and the Aragon province in Spain saw a 2°C (3.6°F) temperature increase in the 20 years between 2004 and 2024 that couldn’t otherwise be explained.“

I can think of one very easy reason why temperatures in these two regions may have gotten warmer, and it's not data centres: climate change.

We know that climate change on average increased temperatures 1-2°c, meaning some areas saw higher increases than others.

As someone else called out "the Bajio region has a metric fucktonne of oil processing in Salamanca and is a plateau region ringed by mountains so yeah"

yeah, I'm sure it was the data centres and uh, definitely not those oil refineries.

In the most extreme case, the researchers said "temperatures increased by 9.1°C", now, unless we're talking about Elon Musk's data centres which are breaking EPA rules by using multiple truck-sized gas turbine generators, maybe there's another reason for this increase?

Let's have a think about where datacenters are built: they tend to be built on greenfield sites, because that is where land is cheapest.

Once one datacenter moves in, other providers usually follow, so we're looking at temperature increase data for land that has gone from literally being a green field with trees, to a concrete, asphalt roads, and metal roofs with very little green space nor tree coverage.

We know the built environment without tree coverage causes the urban heat island effect, where neighborhoods in the same city with and without tree coverage show significantly different temperatures.

"The urban heat island effect can make the country's most populated cities 20 degrees Fahrenheit warmer than nearby areas" – CBS News explaining heat island effect.

@dansup Thanks!! 🥰

BotKit is a TypeScript framework I've been building for creating ActivityPub bots that run as their own independent servers—no platform account needed, no Mastodon or Misskey instance to log into. You write a bot, give it a username, and it federates directly with the rest of the fediverse. It's built on top of Fedify, so the low-level protocol work is handled for you: HTTP Signatures, WebFinger, JSON-LD, all of it. What you're left writing is just the bot logic itself.

Version 0.4.0 is out today, with a PostgreSQL repository for production deployments, a remote follow button on the bot profile page, and a few other additions.

Release notes: https://github.com/fedify-dev/botkit/discussions/20

https://github.com/fedify-dev/botkit/discussions/20

샤라웃되었다. 끼얏호우~

BotKit 0.4.0 is out! This release adds @fedify/botkit-postgres, a PostgreSQL-backed repository for deployments where SQLite isn't enough; a remote follow button on the bot profile page, so visitors can follow directly without manually searching from their own instance; and Session.republishProfile(), which lets you push profile changes to followers without waiting for the next post. It also upgrades the underlying Fedify dependency to 2.1.2, with a few small breaking API changes.

Full release notes:

https://github.com/fedify-dev/botkit/discussions/20

Everyone in tech talks about moats like they're achievements. I keep reading “moat” and thinking: so the idea is to make leaving expensive. Maybe it's just me, but that feels a little grim.

🚀 Spaces has landed on Element X!

Navigate chats by department, project or interest - no more overwhelming room lists.

✨ What’s new: filtered chat lists, dedicated Spaces tab, flexible room creation and previews with context.

👉 Download Element X and explore Spaces today!

https://element.io/blog/spaces-has-landed-on-element-x/

After a pretty long break, I'm back to working on @botkit. The upcoming v0.4.0 isn't going to be a big release—mainly a major @fedify version bump and some general cleanup—but it felt like time to dust it off.

An excellent "Introduction to Writing Systems & Unicode" and its "Large character sets", by Richard Ishida @ri

https://r12a.github.io/scripts/tutorial/part2

#Unicode #WritingSystems #CharacterSets #r12a

《標準國語大辭典(표준국어대사전)》 MCP 서버를 만들었습니다.

旣存(기존)에도 《標國大(표국대)》 MCP들이 있긴 한데, 그냥 標題語(표제어)랑 뜻풀이만 덜렁 주는 데다가, 每番(매번) stdict.korean.go.kr 서버에 要請(요청)하는 式(식)으로 作動(작동)해서 레이트 리미트에 걸리더라고요. 제가 만든 건 아예 全體(전체) 辭典(사전) 데이터를 맨 처음에 받은 다음에 그걸 SQLite에 넣고 照會(조회)합니다.

"But I fundamentally disagree with the conclusion.
The proposed solution is denial and isolation: block the crawlers, withdraw from centralized forges like GitHub, make our work inaccessible to AI scrapers, shun those who use these “anti-ethical tools” from our communities. I understand the anger behind it. But I think it misses something important and misreads the historical pattern that has shaped FLOSS itself."

We should reclaim LLMs, not reject them

https://writings.hongminhee.org/2026/01/histomat-foss-llm/

@jnkrtech Oh, “crazy” and “sane” hadn't crossed my mind, but you're right, and those are everywhere in tech. “Sanity check” especially; I use it without thinking. Good point.

惡魔(악마)가 로봇淸掃機(청소기)에 憑依(빙의)하는 荒唐(황당)한 꿈을 꿨다…

@john That’s fascinating—and it makes sense that you'd notice it more sharply than most. I hadn't really thought about how deep it goes beyond the obvious ones. “Vivid” is a good example; I use it all the time without thinking.

@basil Oh, “enlightening” works well in a lot of cases! Though it leans a bit more toward learning something new, so it doesn't always map perfectly.

@BigTittyBimbo That's a good one, though I wonder if “perspective” itself carries some visual baggage?

I've been trying not to use words like “blindly” and “eye-opening.” Using blindness to mean not knowing or not noticing something doesn't sit right with me. But English isn't my first language, so finding replacements is harder than I expected. Sometimes “uncritically” works for “blindly,” but not always. I still haven't found a good casual replacement for “eye-opening.” I don't think people who use these words are bad. I just don't want to use them myself.

@Kaesekuchen Honestly, yes, it's global state… just scoped per thread/task. The advantage over a plain global is that you get isolation across concurrent requests without threading values through every call. The debugging experience with contextvars was rough in my memory, though I haven't used it in a while. Statically typed implicits feel safer to me because they desugar to actual arguments, so the compiler keeps track. The footgun either way is that context-dependent functions proliferate silently and become hard to refactor out.

Reading this also made me realize I've had a soft spot for dynamic scoping/implicits for a long time… probably since I first used @mitsuhiko's Flask, where the request context object was just there without you having to pass it around. Felt like magic, then felt like a footgun, then felt like a reasonable tradeoff again. Python has since put contextvars in the standard library, which is essentially the same idea.

Enjoyed this wiki post by the author of Garnet on effect systems: [[PonderingEffects]]. What I liked is that it doesn't just describe the design space: it's honest about what the author finds confusing or unconvincing, including a skeptical take on algebraic effect handlers specifically. The Lobsters thread is worth reading too; someone points out that what the post calls “effects on data” is already studied under the name coeffects, which was news to me.

@Yohei_Zuho Windows 98！懐かしいですね…

Having fun with #RSS and #ActivityPub @davew
@hongminhee

---> https://diff.rmendes.net/about

Follow the tool on the fedi with @bot

act II
@mkljczk speed runs contributing to korean fediverse

Hmm, Codeberg seems down now?

https://status.codeberg.org/

@mkljczk Haha, thanks for your first contribution!

7 minutes from installing Hollo to my first pull request, why it's always like this