Hacker Newsnew | past | comments | ask | show | jobs | submit | intalentive's commentslogin

Fine for me on iPhone Safari.

>You have more control here than you have even in C: To allocate bytes, you have to call alloc() on a specific kind of allocator

You can write your own allocator in C. You don't have to use malloc.


you could fully redesign libc to be modern and toss out the null terminated string concept and use fat pointers and slices. But at that point why not consider moving onto a more modern language with things like explicit type conversions, modern PEG type grammar, etc.

Technical analysis is a basket of heuristics. Support / resistance / breakout (especially around whole numbers) seems to reflect persistent behavior rooted in human psychology. Look at the heavy buying at the $30 mark here, putting a floor under silver: https://finviz.com/futures_charts.ashx?p=d&t=SI This is a common pattern it can be useful to know.

Kelly’s indignant attitude and commitment to “engineering excellence” suggest a bright future for Zig. It’s good to see the leader of a technical project get angry about mediocrity.

[..] in a product not people. Insulting people is never a solution.

Sometimes people need to be shocked awake. Reality is harsh, and gentle language doesn't change that.

I've spent time in restaurant kitchens around chefs that believed "some people need to be shocked awake".

The people that got yelled at didn't do markedly better after getting yelled at, but they sure had a worse attitude towards their peers and chefs.

None of the chefs I talked to about it had anything better than "that's how it was when I started in kitchens" as actual justification.


The methods for influencing results within an organization exist on a spectrum, and failing to adequately utilize the breadth of that spectrum is always counter-productive.

If you want to measure the language used by the productivity of the desired outcome. I'd encourage you to survey the ratio of comments talking about the problems with github's very broken CI and UX, with how many people expressed an objection to the language and words used in the announcement. Failure to convey ideas with tact and respect, is demonstrably more counter productive.

I assume you'll choose to dismiss those who object as fragile birds... but then you don't really care about the productivity towards the goal then do you? You just want to be ok with being mean because it doesn't bother you.


Why do you consider that a useful metric? Hit dogs holler, after all.

> Why do you consider that a useful metric? Hit dogs holler, after all.

you do...

> The methods for influencing results within an organization exist on a spectrum, and failing to adequately utilize the breadth of that spectrum is always counter-productive.

Or did you have a different expectation for result in mind? The one you thought would be counter-productive without insults.

My assumption was that ark wanted to put support behind codeberg, and encourage others to take a critical look at how bad github has become, and to consider other options. Not rally additional support and defense of github's actions.


I do about what?

I haven’t actually used harsh language with anyone so I’m not sure what your point is. I have been on HN long enough to know that expressions of strong negative emotion are punished here. That says absolutely nothing about the effectiveness of different methods of influence within an organization.

I think if people are rallying to defend GitHub due to some language that ruffled their feathers and not objective technical merit then they have completely lost the plot as engineers.

As far as Andrew’s goals, I think he has been pretty successful within the framework of the attention economy.


I'm talking about the ideas, threads and conversations that are occupying the head space of others.

> then they have completely lost the plot as engineers.

I think most people who would call themselves software engineers have lost the plot of engineering.

That applies equally to those who are blind to the fact that engineering only exists to create stuff for humans. Most engineers are ignorant to the ability to consider the humans they're supposedly build for.

The point is to make shit better, not worse, and not more inhuman.


If you are hitting the dog unprovoked don’t be shocked if it bites you.

It can be true, that a person needs a wake-up call, but it can also be true that the person(s) doing the "shocking" are sadistic, abusive, or psychopaths.

You’re not mining coal, get real. Either use efficient techniques to make people do the intellectual work necessary to achieve whatever goal you have in mind, or you’re just deluding yourself thinking you’re some kind of “reality expert” while being an asshole, meaning they might still do it, but it would be despite your leadership, not because of it.

Why does intellectual work imply that people doing poor work need to be treated like fragile little birds?

Intellectual work requires a bit of creativity (across all the domains I can think of), abuse, of any kind increases stress, stress decreases creativity, ability to problem solve, and resilience (or the ability to endure the difficulty of solving hard problems).

But even if that wasn't true. There's a significant difference between confronting the harshness of reality. And behaving in a way that makes reality suck more. Every human deserves to be treated with dignity, and a base level of respect.

Suggesting that someone is fragile and weak, because they object to being insulted, or object to the careless and needless stripping of dignity and humanity from people is a wild take.


I dont think porting everything over to React...making the site slower, bloated, & buggier is "creativity".

I agree that people should be treated with dignity...but groupthink & herd mentality often strips people of their humanity.

So the criticism is really about culture & abstract attractors...not the individual people who often act rationally within the context of the system.


I started working on srctree 2 years ago because of how awful github has become. I don't think there's much creativity in this trend line... But the question was; "why is insulting people doing intellectual work bad". Not, "do you think the changes at github are creative", but I do think that the changes require a bit of intellectual work, and that no matter how shitty github has become, it's unreasonable to attack people when unprovoked.

Can you only provide clear and direct feedback on poor work by insulting people?

No but I won’t rule it out for the incorrigible

Ok, but that’s still not effective as a leadership course of action. Calling people names might make you feel like a big man inside, but that’s it, it won’t accomplish anything, that’s only for your personal benefit, not the project, not the product and definitely not the team.

Actually if you completely rule out the possibility of harshness then you are giving license to let yourself be walked over and for standards to drop to zero. It might make you feel like a big enlightened man inside to do so, but the proper application of firmness and pressure is absolutely effective in leadership.

Derision is legitimate way to change behavior when other avenues fail.

A reasonable person that's acting maliciously can be reasoned to stop their behavior.

An unreasonable person that's acting in good faith cannot be reasoned to stop their behavior because they are stupid.

If after attempts to reason with the unreasonable fail, it is not an insult or ad hominem attack to explain the person is acting stupidly.


>Insulting people is never a solution.

That can not be absolutely true.


Nothing is absolutely true, but in this case definitely

This news story was read by investors and leadership inside of Microsoft.

That wouldn't have happened if they hadn't derided whatever idiot decision makers thought it was acceptable in the first place.


The assumption is that the stock picks come from insider knowledge gleaned from congressional duties like sitting on committees. Another possibility is that they are a form of off-the-books campaign donation or bribe, and the knowledge comes straight from an insider who wants to influence congressional decision making.

This makes far more sense. "No, no, it isn't a bribe. It's a sure thing!"

Its almost as if private encrypted chat networks are a thing. Oh, look at that, Mets owner, Steve Cohen paid record fine for insider trading using just that.

I'm not here to shill for Steve Cohen, but I believe that many long/short (fundamental) hedge funds use an "onion strategy" to share insider information from analyst to trader. It's probably defamation to write this, but I do think Steve Cohen has seen plenty of insider information in his trading career provided by analysts that work his fund. That said, he was never found guilty of insider trading. Look it up. He paid a huge fine, but did not admit fault and was not found guilty. Now, there are numerous people around him who were found guilty, but not Teflon Steve!

A “pretrained” ResNet could easily have been trained through a supervised signal like ImageNet labels.

“Pretraining” is not a correlate of the learning paradigms, it is a correlate of the “fine-tuning” process.

Also LLM pretraining is unsupervised. Dwarkesh is wrong.


This is the smallest model in the top 100 of HF's MTEB Leaderboard: https://huggingface.co/Mihaiii/Ivysaur

Never used it, can't vouch for it. But it's under 100 MB. The model it's based on, gte-tiny, is only 46 MB.


Claims that ring true:

* Energy physics puts an upper bound on material wealth

* The disparity between notional and material wealth is large and growing

* Notional wealth figures are largely speculative / fictitious

Claims that could be true (if empirically verified):

* Material wealth is decreasing

* Energy is decreasing

* The monetary system will collapse

Claims that have been implied but not demonstrated or argued:

* There is a causal link between decreasing energy supplies and monetary system collapse

Overall it smells like 2004-era peak oil doomerism. I’m not saying it’s wrong, it could just be early. Intrigued but not convinced.


> * Energy physics puts an upper bound on material wealth

Sure, but that's such a high upper bound that it may as well be false from our perspective in 2025. We use a miniscule fraction of the energy we receive from the sun.


We directly use a miniscule fraction, indirect use is quite a bit higher since that is used delivering ecosystem services we depend on.

Then there is the question of how much of that potential we want to turn into waste heat inside the atmosphere, which is more governed by how much radiative cooling we have rather than how much energy is incident or available on the earth.

I think that humanity would be limited by pollution and ecosystem destruction before energy for most human scale material wealth. The bit where it becomes tricky is energy does change what how easily and fast you can do things, which may place enough of a real world limit.


It's still a tiny fraction. Plants are surprisingly inefficient users of sunshine. And our use of plants is less surprisingly inefficient.

Of course more efficient usage isn't necessarily a good thing.


It would seem you are still just thinking about the human controlled portions of the ecosystem and not including all the energy used to drive to water cycle etc.

It’s hard to control the narrative when people can say what they want on the internet, anonymously, without being punished for it.


"I know why you did it. I know you were afraid. Who wouldn't be? War. Terror. Disease. There were a myriad of problems which conspired to corrupt your reason and rob you of your common sense. Fear got the best of you and in your panic, you turned to the now High Chancellor Keir Starmer. He promised you order. He promised you peace. And all he demanded in return was your silent, obedient consent."


Why do you want people punished for saying things?

I guess maybe this was sarcasm. If so, carry on good sir.


> Why do you want people punished for saying things?

I am not the OP, but I interpreted them as suggesting this serves as a good form of censorship while advertised as improving child safety.


Rather, they are saying governments want to control the narrative and anonymous speech impedes that.


Not OP but while I don't seek "punishment", I do seek accountability. I know that might seem like a flowery synonym at best, or an amorphous piece of jargon at worst, but if we are to treat online spaces as public forums, we need to structure these spaces like public forums, which means having consequences for abject lies. The "but who decides" response is a thought-terminating cliche that we need to collectively move past. Until we stop letting the perfect get in the way of the good enough, we will continue to let bad actors dictate the public understanding of technological issues, and of issues more generally (eg: antivax).


The trump administration in the US also frames its crackdown on civil society in terms of "accountability for lies". But I guess its fine when your side does it.


And here is Exhibit A of those responsible for our current state of affairs


I don't see Trump doing this or his Administration. For the first time in years I'm actually not worried about the FBI and what dastardly political maneuverings they are up to. The CIA is still probably pretty bad. Yes, there are a lot of Republicans who are neo-authoritarians who need to be shut down before they ruin open and free society for a pipe dream. It's like you can't win no matter which party is running things because there are always the freaky lunatics who want to limit your freedoms, expand government, and cover for their own horrible misdeeds.


DHS is the one currently expanding its collective intelligence reach into becoming the CIA+FBI for americans.


> I don't see Trump doing this or his Administration.

It's been a hallmark of his Administration, so you not seeing it is...interesting.

> For the first time in years I'm actually not worried about the FBI and what dastardly political maneuverings they are up to.

In the sense of it not being a mystery because it is more naked in both the direction and the specific approach to partisan political abuse, I guess I could see that, but in terms of not being concerned, the only explanation for that is GP’s “But I guess its fine when your side does it.”


Most claims of 'the other side' is lying are themselves lies. It's mostly people just spinning things to suit their own personal biases (without necessarily even realizing that's what they're doing). For instance the vaccine topic is one I did a deep dive on not too long ago when deciding which vaccines to approve for my children. This [1] is essentially the bible of vaccines - it's a massive study across a large sampling of evidence for all major vaccines, carried out by the National Academies of Science. I'll quote them:

----

The vast majority of causality conclusions in the report are that the evidence was inadequate to accept or reject a causal relationship. Some might interpret that to mean either of the following statements:

- Because the committee did not find convincing evidence that the vaccine does cause the adverse event, the vaccine is safe.

- Because the committee did not find convincing evidence that the vaccine does not cause the adverse event, the vaccine is unsafe.

Neither of these interpretations is correct. “Inadequate to accept or reject” means just that—inadequate. If there is evidence in either direction that is suggestive but not sufficiently strong about the causal relationship, it will be reflected in the weight-of-evidence assessments of the epidemiologic or the mechanistic data. However suggestive those assessments might be, in the end the committee concluded that the evidence was inadequate to accept or reject a causal association.

----

The overwhelming majority of the rhetoric around vaccines, including from governmental figures, is doing exactly what they warn against. There's simply a lot of nuance on most of every issue worth discussing, that people often don't want to acknowledge.

[1] - https://www.nationalacademies.org/projects/PHPH-H-08-17-A/pu...


If you want to talk about Covid “Two weeks to slow the spread” was the foundational lie that they told that did more damage than almost any lie I can remember. That is solid truth right there.


> but if we are to treat online spaces as public forums, we need to structure these spaces like public forums, which means having consequences for abject lies. The "but who decides" response is a thought-terminating cliche that we need to collectively move past.

In order to "move past" that, you have to find a way to address official lies and cases where the majority is wrong.

.

For example the official denial of the fact that the Wuhan lab was researching things similar to covid-19. (Doesn't matter whether it actually came from there.)

Or the official lies about mask effectiveness. (Regardless of whether they're effective or not, the government told people things that it believed at the time were false.)

Or the lies about the world's best anti-parasite medication (that just isn't an antiviral) being dangerous horse-paste.

Or the lies about Hunter Biden's laptop being Russian disinformation.

Or that still-ongoing culture war topic where both sides claim the other is lying.


It'll be alright. We've dealt with manic authoritarians who dream of planetary control before. Just another quick world war and the development of an even more sinister superweapon and we'll be right back to thinking in sane, evenhanded terms. Or dead.

Or, you know, we could huck our failed systems out with the trash instead. Reinvent democracy to be more direct and flexible. Could be nice.


That's not exactly true. Social engineering at scale is becoming better and better. You don't need to remove dangerous opinions, you simply need to make sure nobody cares about them. And the fun thing is, most people say that they support freedom of speech, but when they see an opinion they dislike, they say the opinion should be blocked. This is why all the laws restricting freedom of speech pass without major issues.


IBM and Nvidia speech to text models are also SOTA (according to HF leaderboard) and relatively lightweight. Replicate hosts those too, although some (like Parakeet) run easily on consumer GPU.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: