I grew up before the Internet was mainstream and I don’t remember this. We all had access to basically the same information and some of us still had worse or better ideas than our peers. Access was always only one part of the equation; beyond that, you need the information to be useful and accurate (big problem on the Internet), you need the desire to engage with that information, the ability to process and understand it correctly, the ability to discern when factual information is being cherry-picked or otherwise used in misleading ways …

If you trip over on any of those points or whatever else I’ve forgotten to mention, you come out the other end with bad information, access be damned.

Not even fake, here’s a news story from 15 years ago featuring the same image: https://phys.org/news/2010-12-million-years-oxygen-drove-evolution.html

That’s eight years before Among Us was released. I checked archive.org as well because … well, it’s sus. But they have captures there too, it’s not a backdated modern post.

It does, yeah. I’m not sure if there’s any advantages to Floorp’s implementation or anything else that makes it preferable to upstream or a more shallow fork.

Yeah, it’s a Firefox floork where the main differentiator is vertical tabs, IIRC.

The last subheading in the article is “Who cares?”, by which they mean “Which chimpanzees within the social group are responsible for providing medical care?” I didn’t notice the final section initially, so I thought I had reached the end of the article then there was just an exasperated large print:

Who cares?

You could try something like a network filter that is out of the control of the user (e.g. on the router or something like a Raspberry Pi running Pihole), but you’d probably have to curate the blocklist manually, unless somebody else has published an anti-LLM list somewhere. And of course, it will only be as effective as the user’s ability to route around that blocklist dictates.

LLMs can also be run locally, so blocking all known network services that provide access still won’t prevent a dedicated user talking to an AI.

In a two-party system like the US, assuming all supporters of a particular candidate share their ideals is folly. There’s a major difference between being someone who actively describes themself as a “moderate” or “centrist” and somebody who just happens to be a voter in a flawed democracy.

I like how the author says Google killed the Fitbit Sense so they could sell an inferior product without having to compete … then said author reluctantly buys and recommends that inferior product. I’m way pettier than that. I’d use anything else, even if it sucked, rather than directly reward a company for fucking me over.

That’s closer but rather than being a wrapper, it takes the original architecture’s instructions (MIPS in the case of N64) and generates a C/C++ function which implements that instruction. Then you call those functions in the same sequence as the original compiled machine code ran instructions.

That’s a relatively inefficient way to make a port, because you’re basically reimplementing the original CPU in software, hence why some have described it as emulation. At the same time though, most recompiled games are like 15-20 years old, so a bit of overhead on a modern PC isn’t going to hurt you too much.

But anyway, unlike WINE, the original binary is not used any more after recompilation. Instead, you have a native binary for the target platform, the translation having occurred at the time of recompilation (when you built the port binary).

Not really. The Ship of Harkinian ports are based on decompilations, which is where you reverse engineer some equivalent source code using the final binary as a reference point. Then, you can port that source code to anything else you can build for, like a PC, phone, Wii U or Dreamcast.

Recompilation, which is what this project is, is closer to (and some have gone as far as to say that it is) emulation. It’s taking the final binary and then, without actually working backward to get source code, translating the raw instructions directly into code that compiles for a different platform.

It’s kind of difficult to get across the difference without being familiar with what both are doing behind the scenes, because the result is obviously similar. Both require human intervention, but decompilation is the more labor-intensive approach, while recompilation is somewhat more automated.

The advantage of former is that you end up with a relatively human-readable codebase to work with, while the latter doesn’t bring you any closer to understanding how the game works internally. Both ultimately allow for porting the game to new platforms. Decompilation will almost certainly result in a more optimized final game, because it avoids the overhead of “emulating” the original architecture. However, for the same reason, recompilation can be generalized to other games that originally ran on the same hardware.

Ah, the learned helplessness<->weaponized incompetence spectrum.

My answer is current era regardless, but do we keep our memories and go back, or is it as if we were born in that era? If you went back 500 years with the knowledge that the Super Nintendo and the Internet exist (the two inventions we have that they didn’t have in the 1500s), that would be unpleasant. But if you didn’t know that and were accustomed to getting your entertainment from court jesters and public hangings, I guess that would be slightly less awful.

Like everybody else has said, there’s a lot of things we have now (by which I mean two) that are better than anything there was 500 years ago, even for monarchs. Regardless of whether I knew about those things in monarch form, the version of me that’s making the decision knows, so … nah.

JERRY!

alive and somewhat kicking
freenode

Haha. I think you’re a bit out of the loop, Freenode “died” like four years ago. Some fucking weirdo who claims to be the rightful king of North and South Korea bought it, started selling advertising on the web site and tried to turn the IRC network into a “cybernation” with himself as ruler. Cool, right?

The people who actually ran Freenode all left to found Libera.Chat and basically everybody moved to either there or the OFTC. Meanwhile, original Freenode was shut down entirely (i.e. all users, services, etc. dropped) and booted from scratch as … whatever the fuck it is now, some coked-up royalty LARP.

What’s the quality like of the people who are still on Twitter in 2025? Does the fediverse want them? (These are real questions, I have no idea if there’s still any decent people on Twitter.)

Jeffrey Toobin was pretty good at it.

• ABC iView v4.16.1 [4148], the last version of the Australian streaming service before they started requiring a login, IIRC this doesn’t even work any more so I just stopped using it
• the legacy version of Discord, more specifically the Aliucord mod which backports some modern features along with a bunch of optional plugins
• Simple Solitaire Collection, an open-source card game collection; the developer took it closed-source and ad-supported so I just stopped updating
• Skype on the last version before they added Copilot, but with Skype shutting down that’s not really useful information to anybody

What are your favorite search engine alternatives? Ideally no ads or favored content, and with various useful filters

There’s the obvious suggestions like DuckDuckGo, but if you’re used to the (actual) results you get from Google, it’s worth checking out Mullvad Leta. This is a search engine that proxies the results of either Brave or Google, so you get their results without them tracking or advertising to you. This is the only way I use Google when doing text searches these days. Sometimes I still have to break it out to do image searches and stuff, as Leta only handles standard text, but it’s been handy as a way to divest from Google without giving up the benefits.