Groxx

Or I just disagree with your premise (there are lots of Bad Influences from chronological sort, and I'm far from the only person who considers it "an algorithm" in the social media sense because of that), and you're looking for reasons to dismiss things you disagree with.

Fair (I haven't been using the encrypted-reasoning systems, though this is common in open ones - I'm kinda surprised it's an option in encrypted ones too), though what they're doing here is cross-user replays in addition to cross-model.

>We take a trace produced by a frontier model, replay it into a weaker sibling, jailbreak the weaker model, ...

Ha! I've been wondering if replaying across models would work, ever since https://blog.cryptographyengineering.com/2026/05/29/fooling-...

I'm honestly rather curious if this was intentionally allowed, it's the sort of validation that's easy to miss (particularly if you're wading into the vibe waters). Seems like something that'd be absolutely riddled with possibilities for shenanigans.

The "Ship in Distress" title editorializing here is part of a very active flood of "lol zuck broke the law" both here and elsewhere.

Which is probably why it's now flagged off the front page.

It's being drummed up to "the ship was in distress and this is illegal under maritime law" absolutely everywhere that I've seen.

Like this post. The "Ship in Distress" part of "Mark Zuckerberg's Yacht Refused to Help Ship in Distress" in the display text here is inaccurate, in a legally-significant way. (Was this the original headline text? I don't see any evidence of it, at least)

Or comments here, e.g. https://news.ycombinator.com/item?id=49251495

Yeah, drag-scrolling and page-jumping on eink is solidly in the "not great" category.

Maybe you could draw a line showing where [some line of text] is moving to? Like starting from where the finger starts, draw a "timeline" of locations as they move their finger, and when they release you draw the whole page:

    ┌───────────────┐ 
    │               │ 
    │         drag  │ 
    │          ┌┐ │ │ 
    │ │        └┘ │ │ 
    │ └┐          │ │ 
    │  └┐         │ │ 
    │   └┐        │ │ 
    │    │     ┌┐ │ │ 
    │    └─    └┘ ▼ │ 
    │               │ 
    │               │ 
    └───────────────┘ 

(wow, lol, mobile Firefox renders that horribly. I wonder if I'm missing a font...)

You can do small-bounding-box "draw black on top" extremely quickly, it's how reMarkable tablets get their near-20ms pen-to-display latency.

I'd probably also like a "this is where the previous top/bottom/where you dragged is now" temporary underline across the screen at the end, for easier visibility. Just clear it a second later.

Interesting - that definitely seems like a more useful approach than "click to advance one page in that direction, but only if you don't click on the bar itself". I kinda wonder if that's possible to do OS wide nowadays...

From the original article they link to: https://alaskabeacon.com/briefs/cruise-ship-helps-stranded-s...

>According to the U.S. Coast Guard and the Marine Exchange of Alaska, which monitors ship traffic in Alaska, there was no distress call — just a general request for assistance.

As much as I wish bad things on Zuckerberg, this seems pretty clearly on the "rude but not illegal" side of things. Super-yachts are kinda by definition rude already, so this seems entirely expected.

> (3) the user expressly and unambiguously requested the specific media, media by the author, creator, or poster of media the user has subscribed to, or media shared by users to a page or group the user has subscribed to, provided that the media is not recommended, selected, or prioritized for display based, in whole or in part, on other information that is not permissible under this definition;

incoming wave of "follow back"-encouraging UX patterns, "subscription packs", and "you might like [add all]" buttons in 3... 2... 1...

Someone chose "most recent" when implementing "most recent", and then that decision automatically controls what is shown next. They could have chosen "most recently liked" or "most liked" or "minimum MD5 checksum" or "alphabetical" or "most liked in past 24h" or "likes / hours * followers / followee_ratio * poster_birthdate_days_diff(admin_birthday) * tfidf_rank_matching_books_published_in_1885" - all of those would be algorithmic sorts in my book. "Most recent" is just one (among multiple) of the simplest ones to predict the behavior of.

A fully human-curated feed would not be "algorithmic", in contrast. Print newspapers do this, and sometimes online ones to varying degrees, and most blogs. You could even let people build their own (e.g. "boost to my popular feed [at position X/newest/oldest]"). Essentially none do this though, they all choose algorithmic ones of varying complexity - it doesn't make them non-algorithmic to choose a simpler one.

It'd need some kind of compile-time hardware-feature-detection, yea? That seems probably feasible since proc macros can do essentially anything they like (worryingly).

There's a "trending" feed on nearly every implementation including the official Mastodon one (hashtags have their own "trending" too), and chronological definitely fits "an algorithm" imo - it has all the hallmark problems of one, such as encouraging bad behavior (frequent spamming is the most effective strategy to be seen). Lemmy is even further down this path, with "hot", "controversial", "active", etc, and defaults to "algorithmic" sorts for its front page and all comments.

Fediverse absolutely has "algorithms". People generally want them. I agree that it's broadly less user-hostile (I'm very much a fan), but it's very much moving down the same paths as twitter (mastodon), reddit (lemmy), etc.

> 10. Line-art or halftones over shade gradients (images).

This is particularly true for eink, because black <-> white is dramatically faster and more reliable (less ghosting) than anything involving any gray. It's why basically all of the "fastest" modes are pure black and white, and why UI shapes are generally pure black and white (though text might have grays).

A lot of that is simply formula authors / application devs who don't know what they're doing (python@3.10 and other versions are a thing, and have been for quite a while now, but they're not always used and devs don't always keep track of the version they need) and people not updating their software for years (pythons are on a 5 year cycle everywhere, homebrew included: https://devguide.python.org/versions/ and https://formulae.brew.sh/formula/python@3.10 ).

Python in particular is well known to not be a stable target. For anyone. By design. If you expect long term use of a specific version of code, use a different language. It is not at all homebrew's fault that they're how many people discover that.

Modular ports are a physical disconnect between "thing that routinely experiences stresses and has a relatively short expected lifespan under heavy use" and "extremely expensive and long-lived motherboard", usually a separate circuit board connected by a ribbon cable. They serve a completely different purpose from a dongle that converts a common port to a less common one.

Framework laptops have another decent strategy, where you've got on-motherboard ports that have relatively few usage cycles, and are protected from most forces because the easily-replaceable custom-dongle-squares are held in place quite securely. They take the damage, protecting the motherboard. And as long as the motherboard ports are still good, you can even forego the modular thing and just plug in directly if needed.

Red Blob Games has quite a few S-tier posts, highly recommend exploring further if this is at all interesting to you

Instagram is even easier if you stay logged out. Sometimes you're just blocked from doing or seeing anything except one page of thumbnails.

(instagram is very likely the absolute worst frequently-used site you could choose for your business's hosting)

We've got laptops with modular USB-C ports that solve this effectively perfectly. An audio one would be even simpler. The only reason you don't see it everywhere is because they're being outrageously hostile to repairs, and choosing to charge you hundreds of dollars for repairs on a part that they know is easily broken.

There were also quite a lot of examples where AMP pages were slower after the shift, but got ranked significantly higher. It definitely wasn't about speed, they already have that project and they could just listen to it.

AMP was just one of many smash-and-grab attempts to steal the internet, wrapped in a fake-gold-encrusted PR-infused box. Google has quite a lot of them.

"You must enable DRM to play some audio or video on this page" pops up in the strangest places...

The vast majority of people I know who have done phone/chat support have had a spreadsheet (or fancier tool) of common phrases and mandatory boilerplate answers, an overview of the company they're doing it for (sometimes just a couple pages in a word doc), and tone/tool training at a minimum. And generally that tool has a tree of common steps that are also generally mandatory, because they punish rather harshly for going off-script. It's not much, and it gets you the sort of support that people often think of with ticketing systems: impersonal and inaccurate, favoring the company.

But it's rarely this inaccurate. Or if it is (e.g. missing a major product launch), it's fixed in a day or two.

Yea, it's terse and clipped in some sentences, and then changes tone abruptly and randomly, and paragraphs really don't flow together at all. It feels awkward to read, and there is a lot more to read here than there needs to be...

If you have no training or knowledge-base to search, sure. But then you'd be an awful support-team employer.

Unfortunately agreed. Browser sandboxing is usually quite a lot better than apps. The table they provide is rather biased though, many of the things they show in PWA's favor are only true in the most trivial sense, and are easily worked around, both technically and widely used in practice.

Also browsers can use extensions to modify pages, and that opens up SO many possibilities. And bookmarks. And history. And tabs. And search. Most apps never implement any of these, much less all.

I do wish we had a more widely supported system for push notifications though. WebPush exists and broadly works quite well, but provides nowhere near the control, reliability, and observability that native apps do. ...which is a good thing when you don't want them, but crippling when you do, and it contributes to the general "few sites do it at all" problem. E.g. Slack doesn't use it, nor does Zulip (particularly painful because their mobile web UI is much better than their app), nor most chat apps in general even though they're an ideal scenario.

Some hotkeys are hard or unusable (e.g. some of the ones that conflict with the browser itself, like esc and ctrl-l and ctrl-t/w/q). Plenty work just fine.

On an 8GB GPU and 32GB laptop: ~5 words/s while running in Qubes via ollama with completely default settings (I don't have an install at the moment that'll tell me tokens/s). Not exactly a highly tuned setup, but it's a ballpark at least :)

Tolerable and usable for some things, though thinking makes it take about a minute to reply in many cases. But getting this kind of thing to run on 8GB of VRAM is the main benefit of the mix-of-experts setup: it can do partial GPU loading and get a ton better throughput than a similarly-sized dense model (like 5-10x, sometimes more).

Yep. I consider it essentially "explaining it to myself". It serves a similar purpose as trying to teach someone else - trying to write something concisely and clearly is difficult unless you actually understand it, and then it often feels obvious. So if I'm struggling, I should probably think about it more (which the struggling is also accomplishing), and maybe go back and reread it to make sure I get it.

It's kinda the purpose of note-taking imo. At its core, it's a mechanical way to enforce "think more about it". Simply putting words down in some form is at least partly achieving that, though if it's mindless it doesn't tend to provide very much benefit beyond that point.

(this is also why I despise mandatory note-taking in classes. they have universally expected something like a dictation, and that isn't how I take notes. (I'm fully aware these are just awful teachers and it isn't how all others operate, and there were many other signs they were awful))

Ribbons are splotchy and weird, and have designs that don't line up and stars that are misformed.

Are they going to use it as an official poster or something? It is it just a side competition? Feels like they might come to regret spreading it. Both the first and the second smell rather strongly of AI, though the winner seems the most egregious.

Yea, same. I had many opportunities in college (who doesn't?), never did. It's not hard, you just need to actually try to learn.

If I was in a course where it was so poorly taught that cheating was the only feasible way to pass, I'd go talk to the dean. Things like that don't get fixed unless people complain, because how would anyone making decisions know?

It sometimes works, there's no reason to leave stuff like that for the next group of people to suffer through too. E.g. I'm extremely proud of my wife for this kind of thing, as she managed to get nearly a whole class to complain about an awful pseudo-science-pushing teacher, and got them kicked out in fairly short order.

Along somewhat similar lines, my main division between the two is that sci-fi generally touches on the interaction between "people" (often the main characters, but not usually limited to them) and the world's systems (politics, war, how future tech changes society, etc), while fantasy generally focuses on individual's internal struggles (how they deal with the violence in their swords and magic world, how they get along with their squad, how they deal with the politics they're stuck in).

Which isn't to say one is more personal than the other, just that the focus is different. How the world as a whole handles something can easily be covered with a single character's thoughts, and they can be worried about how they interact with the world that surrounds them.

No division is perfect of course, but I think it fits pretty well with a lot of the genre-bending works, and better predicts who will like what book. Some are totally in X genre because of the setting itself, but more often I find people sticking to one or the other because they like that kind of story.

(Also Le Guin is fantastic, very obviously an exceptional author, and you should absolutely read some of her works)