ezst

Curiosity is a teenager now? Damn, I didn't need to feel this much older today..

People having extremely opposite user experiences with LLMs. How could this be?

Yes I wonder why. Teen pregnancy sounds so amazing and inconsequential for individuals and societies alike that teens should just be inoculated as a rite of passage or something...

Sounds good, until you start arbitrarily punishing people who, for one reason or another, just can't have kids (reasons can be many, for one, biology is a bitch).

Some people prefer organic grown food for all kinds of reasons, does it matter to you they would want the same for code? (Also, I'm not picking a side here)

And with dnf/rpm updates being transactional and easy to undo with `dnf history` or `dnf distro-sync`, I never saw the appeal for a day-to-day OS of fedora atomic & al.: only a worse user experience with a slower system, slower updates, and terrible disk usage in a time of storage scarcity. I keep missing the obvious and telling myself that I'm dumb for it, but OTOH, this box is running fedora 44 with some big COPR (the latest one being for plasma 6.7) admirably well, and the most admirable of all is: this OS was initially installed as fedora 27 and incrementally updated flawlessly for almost a decade now.

I mean, this reminds me of the early imagenet days, when people were first trying to explain the unreasonable efficiency of neural networks at interpreting content in images. They found out that, after enough backprop, some layers of the network became specialized in contours extraction, shape extraction, etc, in ways that can be seen as analogous to earlier CV techniques (Canny/Hough transforms, ...).

Later, Google was having fun feeding whole bunch of youtube content to artificial neural networks, unsupervised, and figured that certain parts of the network would, too, specialize, only to have the activation functions be run backwards and render an abstract image of a cat¹.

None of that is terribly new or surprising for anyone having studied and dealt with neural networks. The only difference today is that the field has completely flip flopped from approaching the subject with scientific rigor and cautious excitement to being a clueless billionaire infinite money printing machine fed on deceiving anthropomorphism and FUD.

¹: https://blog.google/innovation-and-ai/products/using-large-s...

Pretty sure there are already a dozen vibecoded proxies for HN doing just that

> that makes me feel I'm not talking to a self sufficient entity.

I've seen this happening a lot in the recent times: people who are generally not very good at their job tend to offload disproportionately more to LLMs, and it's so damn annoying that their incompetence now comes sugarcoated in lengthy LLM babble for the sake of desperately trying to sound convincing. This is wasting me high single digit hours every week, not to mention the frustration of battling an asymmetrical fight: it takes them seconds to produce something that will take me minutes to read and hours to react upon. This needs to stop.

Edit: typo -le+me

I find GBoard to remain superior in multilingual typing, futo just can't seem to be able to switch to other languages consistently, even with the multilingual option enabled,

it's also not as good at "recovering" from typing too many letters (Gboard sometimes adequately completes with likely shorter words)

If you haven't already, you might want to read the twitter thread at the root of this HN discussion. It essentially postulates that all Chinese 3D Printer manufacturers are incentivised to follow in the steps of BambuLabs (being members of the state apparatus and subject to the same rules), while suffocating the competition outside of China (by benefitting from unfair advantages under the umbrella of a nationalistic development plan).

I don't want to sound like a Prusa shill or anything, but if they see Chinese state-sponsored competition as an existential threat (and I see no reason to doubt either claim), then I imagine that, from Prusa's perspective, it's a losing proposition either way ("damned if you do, damned if you don't"): anything they develop in the open directly strengthens their competitor and doesn't elevate the playing field for the benefit of all (unpunished licenses violations) ; anything they keep to themselves turns Prusa further and further away from their ideological stance, more akin to their enemy and less relevant.

> the effort has been stalled for years

All because of some cargo-culted command!

For a while (a decade+), I was running CentOS on my servers on the same assumption of long time stability and ensuing peace of mind. Then I figured that over such durations, the ecosystem drift becomes significant and keeping applications up to date and running on top of the OS becomes an increasing challenge (with the more "infrastructure" packages like glibc, python/Apache combos, GCC, ... slowly becoming incompatible with the latest applicative stack).

Then I figured that version upgrades were miserable, not just because I had painted myself in a weird corner with ungodly packages mix-ups, but because the upgrade path was always best-effort. I think I gave up during the 6 to 7 transition, as I realised that all I needed was fedora: with yearly or half-yearly updates I have no need to fight the distro's packages: stuff stays current and in working order, major distro upgrades go smoothly, downtime is minimal. I'm not considering going back to any "server distribution" ever.

When the field is so crowded with astroturf and myopic takes that you can no longer tell what's sarcasm and not.

I mean, that's probably what it is, a last squeeze while something somewhat remains of their Plex brand image, knowing the war to Jellyfin is over and lost. At that rate, Jellyfin could start a commercial entity, partner up with mullvad or whoever, and offer a STUN/TURN proxy service that they could charge cents/dollars a year and still make bank.

> The bookmarks feature which is supposed to be the solution for short-lived branches is hard to understand though.

yup, it came up as a response to git and people getting more and more re-wired into thinking of "branches as pointers", with its own bag of implicit state and behaviour. No wonder it felt clunky and unpolished used in this context. That said, it was a pretty effective way to globally distribute a namespace of string→commitID pairs, i.e. a list of commit bookmarks, and at least in that sense the feature wasn't a total misnomer/let-down :-)

Both things can be true (the second being that staging was never necessarily not a desirable abstraction in light of easily and safely amendable commits)

Then you made the right choice, and it's only unfortunate that not enough of you made the same to tip the scales in favour of mercurial, or at least to avoid the monopolistic situation we are in now.

I might say that I stopped thinking too hard about this many years ago: now I use hg-git whenever I must interact with git, but do regret that the more sophisticated/qol features of hg (branches, evolve, phases, topics, ...) don't exist in git, and that's the real bummer: the more limiting alternative won, then stagnated, and we are all worse off as a result.

Do they? They came to the realisation that they control a sufficient fraction of the market that your preference as a consumer no longer matters to them.

> select a rectangle and perform OCR

You get that out of the box on an up to date KDE, from the screenshoting application Spectacle.

https://www.scala-js.org/ is quite phenomenal.

> never use code that can stop working until you pay ransom

Funny because that describes pretty much exactly "cloud-first" software architecture, and people jumping on it in troves, unexplainably.

ok, now bring-up a photo and datasheet of the flagship model from each of those brands, and articulate how they effectively depart from one another on metrics such as ergonomics, operating system, and hardware specs such as CPU/GPU/modem.

- The ergonomics story is: all those devices look and handle the same

- operating system is: a duopoly between Android and iOS

- chips: a triopoly between Qualcomm, Samsung and Apple

those manufacturers are not innovating as much as copying each-other's formula, using a very short list of identical suppliers, with no room for error. If that's not the definition of consolidation, please show it to me.

> This should be for the market to decide, not EU bureaucrats.

Ohh sweet summer child... We are in an era of obscene consolidation, in pretty much every sector, wealth is being consolidated to degrees unseen before, oligopolies enshrine their dominance via regulatory capture and a plethora of unfair practices. There's just no competition left to suggest that "markets can decide" of anything beneficial for our skinny bottom lines..

Not sure I agree. You can "blind type" on a physical keyboard, and even if it has less sophistication in the way of inputting large amounts of text (lack of auto complete, lack of fuzzy typing/auto correct), a calculator is purpose built with tons of shortcuts and contextual menus that you access from muscle memory without second guessing yourself. Right now, if I've got a mildly complicated mathematical expression to type, I'd rather do it on a last-century calculator rather than e.g. on Android's GeoGebra.

> MacOS by far has the least bugs and kinda just works.

MacOS is sometimes so weird and inconsistent that it's hard to tell whether it's a bug of Apple usual's "you are not smart/cool-enough to understand" kind of feature.

> But the sad reality is that x86 processors are falling behind in performance. The Intel Core Ultra X7 358H is a whopping 35% slower than the M5 Pro

35% is the generational difference between the M4 pro to the M5 pro¹. Don't drink Apple marketing koolaid: this has less to do with x86 falling behind than it has to do with Apple using their stash of gold to outbid the intel/AMD competition out of the latest TSMC capacity.

An M4 Pro 12c gets 32731 out of TSMC's 2nm-E².

An AMD Ryzen AI 9 HX 370 gets 35093 out of TSMC's 4nm³.

The true unsung hero of the "Apple M miracle" is TSMC, not ARM, and Apple mostly in the sense that it has the deepest pockets.

With the first M chips, anyone who could afford to wait 18-24 months was pretty much where Apple was at. This decreased to 12-18 months in the last couple years. Panther lake signifies that it could further decrease to 9-12 months.

¹: https://www.cpubenchmark.net/compare/6346vs6345vs7230vs6397v...

²: https://www.notebookcheck.net/Apple-M4-Pro-12-cores-Processo...

³: https://www.techpowerup.com/cpu-specs/ryzen-ai-9-hx-370.c365...

I happen to have a pair of airpod pro 2 and some Sony's (the airpods being a gift). Nothing about the airpods strikes me as being clearly superior, which is surprisingly rare for Apple products: they tend to all stand out in one way or another, whether that's essential to the experience or not (I have no use for an iPad, but I can tell a good display when I see one, I wouldn't keep a MacBook if one was gifted to me, but I can appreciate a good trackpad, …).

Airpods? I have nothing bad to say about them, but nothing good either (I rather take the Sony's with better NC and battery life on a flight, better audio quality and painless equalizer).

> Or is it more likely that some consumers judge the utility differently from you?

That's possibly the case, so help me, what I am missing?

> Siri is STILL utter garbage.

Always was. If Apple had an easy trick to make it any good or useful, they would have already. LLMs aren't that, so Siri remains what it is.