atq2119

This doesn't work across forks, though, and therefore doesn't compose with GitHub's model where you push your changes to your private fork and then open a PR from there.

There's also no native UI support for it in GitHub -- I'd expect to have a navigation element for stacked PRs like in Gerrit.

It's not the same price. The shipping companies charge you a processing fee in addition to the actual tax, so you often end up paying a lot more. Bulk shipping is more efficient.

> We do have farmers growing apples, but there's no harvesting those during our snowy winters so our tariffs on apples are just active when local apples are in season

I'm curious: apples are fairly easy to store in bulk. That's a large part of why we have them year round.

So what's stopping someone from importing apples outside the season and then sorting them to circumvent those tariffs?

I don't think the answers from others in this thread are particularly satisfying either, at the time I'm writing this.

Thinking about it a little bit, I'd say the major challenge for converting from shared to static library is that shared libraries have resolved relocations from text to other segments.

In order to create a static library, you need to make it so that text and data can be moved relative to each other again.

At a minimum, this requires disassembling the text segment to find PC-relative addresses and recreate relocations for them. That doesn't sound impossible, but there may be pitfalls I'm not thinking of right now that make it hard to find all necessary fix ups.

If buying hyper luxury goods was the only thing that this wealth was good for, I'd agree.

But it's not: many of those billionaires are using their wealth to subvert democracy. They are quite literally enemies of the state.

The only lasting defense is to prevent this level of individual wealth accumulation.

Diffusion could potentially be more efficient for local inference. With auto-regressive models, token generation is basically one token at a time, and so is not compute intensive at all -- it's bandwidth bound. With diffusion, you always run the model on a decently sized batch of tokens, so you should be (close to) compute bound even for local inference.

If the "output quality per compute" is roughly the same for diffusion and auto-regression (is it? I have no idea...), then diffusion will be much more efficient for local inference because the same amount of compute can be packed into a much shorter time period.

> Keep in mind that what the fed article is calling a “saving gap” is really more of a fiscal gap that been plugged by US Treasuries.

They're two sides of the same coin, but the levers of control and causality aren't symmetric.

In particular, the US government doesn't have direct control over the savings behaviour of anybody, especially not of people outside the US.

That's why most policies that aim to reduce the US government deficit are bound to either fail or have undesirable negative effect elsewhere.

Status quo bias. The previous behaviour was also affected by government policies including taxation and infrastructure spending.

There is no objectively neutral baseline of preferences here as long as civilisation exists.

Also beware the Fundamental Attribution Error.

CMake can't do everything autotools does, but the stuff autotools does which CMake doesn't isn't relevant anymore in today's world.

The fundamental curse of build systems is that they are inherently complex beasts that hardly anybody has to work with full-time, and so hardly anybody learns them to the necessary level of detail.

The only way out of this is to simplify the problem space. Sometimes for real (by reducing the number of operating systems and CPU architectures that are relevant -- e.g. CMake vs. Autotools) and sometimes by somewhat artificially restricting yourself to a specific niche (e.g. Cargo).

I'm not particularly fond of the whole "privilege" discourse, but this comment is a great example of somebody completely failing to understand a privilege they have. Which they share with many other people, sure,[0] but there are many people who, through no fault of their own, do need to worry about others learning about their location.

[0] Which is probably one reason why the discourse grates some. Privilege still sounds to me like it's something exclusive, like a 0.1%er thing. Naming stuff is hard.

Not generally, no. This is true for some chips, especially (very) old or simple cores, but it's not something to lean on for modern high end cores.

Possibly somebody downvoted because "thank you" in all caps is not a substantial contribution to discussion. It feels like the kind of low effort stuff you'd see on reddit.

Also, commenting on downvotes is generally frowned upon.

See also: code review

Ironically, that last point has come true!

(Though to be fair, the button tends to be labeled both start and stop)

Has that really been established? The observable universe is finite, yes, but I wouldn't think that automatically implied that the universe as a whole is.

It doesn't, but whoever sells it needs to contribute back whatever they did with it is almost as good.

It's not quite as good because some games still get griefers who sell versions on some market places without otherwise giving back to the community who maintains them.

But it's better than MIT for sure.

A lot of compensation works that way, to be honest. First order is that you get a percentage of whatever river of money you sit close to, regardless of effort or skill.

Calling it a red flag may have been too harsh. It's certainly not an immediate no.

However, like it or not, it is a signal because it means you deviate significantly from the mode of the distribution. And a sober application of Bayes suggests that if anything, all else equal that signal is a negative one.

Wearing a suit to a technical interview is an immediate red flag. Everybody knows you don't wear suits in this industry, so what's your motive? Your ability to wear a suit is irrelevant for the job, so what weaknesses that are relevant are you rather clumsily trying to hide?

A couple of things. In practice, most goods are actually priced as cost-plus, with very thin margins, and any response in prices is highly asymmetric.

If the purchase costs for a good suddenly increase, retailers will increase prices very quickly because they would otherwise start losing money very quickly.

If demand for goods decreases, that doesn't affect the retailer's cost of goods at all, and if they were to reduce their prices, they'd eliminate their thin margins. From their perspective, it's better to sit on inventory for a while.

So any downward pressure on prices would happen much more slowly.

BTW, this kind of asymmetry is why a bit of inflation is good, actually. Inflation acts as a universal and permanent downward pressure on (real) prices, in the sense that if retailers and others are unable to justify an increase in nominal prices to their customers, their real prices will drop.

I actually agree with you, but the point is the balance.

Don't steal. Don't share embarrassing or humiliating information you may come across.

At the same time, there should be safety from prosecution overreach.

I ask for this mostly not for my current self but for "kids" (including young adults, e.g. college students) who are on a hacker journey in the original sense of the word. As a society, we should encourage rather than stifle that sort of exploration.

The interesting bit is that social expectations matter.

There is a social expectation that people can generally only enter your home with explicit permission, and so if they didn't invite you it's trespassing even if the door is unlocked. But maybe you have some close friends who you get used to coming over and just entering even if you may be out at the moment -- and then it's not trespassing anymore.

Remote computer access is a much younger phenomenon than people living in houses, and so social expectations aren't as established. There's a legitimate need for discussion there.

For example, if you have an open webserver that you want people to access, is it trespassing if people fiddle a little with the URLs and encounter documents that you didn't mean to put out there? I'd argue it would make for a healthier and more tech-savvy society if we didn't consider that trespassing.

If we try to push the houses analogy further, it's a bit like inviting people into your house for a big party, and then somebody enters a room that you didn't want them to enter. It's a faux-pas, but you'd probably also have a hard time if you tried to label it trespassing.

Maybe? I'd love to hear if there are some good tools for it that can be integrated into typical setups with Git repositories, Jenkins or GitHub Actions, etc.

I agree with your assessment of how stupid this is, but I'm not surprised.

To be clear, there are good reasons for this different mode. The fuck-up is not testing it properly.

These kinds of modes can be tested properly in various ways, e.g. by having an override switch that forces the chosen mode to be used all the time instead of using the default heuristics for switching between modes. And then you run your test suite in that configuration in addition to the default configuration.

The challenge is that you have now at least doubled the time it takes to run all your tests. And with this kind of project (like a compiler), there are usually multiple switches of this kind, so you very quickly get into combinatorial explosion where even a company like Google falls far short of the resources it would require to run all the tests. (Consider how many -f flags GCC has... there aren't enough physical resources to run any test suite against all combinations.)

The solution I'd love to see is stochastic testing. Instead of (or, more realistically, in addition to) a single fixed test suite that runs on every check-in and/or daily, you have an ongoing testing process that continuously tests your main branch against randomly sampled (test, config) pairs from the space of { test suite } x { configuration space }. Ideally combine it with an automatic bisector which, whenever a failure is found, goes back to an older version to see if the failure is a recent regression and identifies the regression point if so.

Why would a 25% drop in energy use necessarily indicate an energy crisis?

Isn't it far more likely that a shift in economic composition changed the energy consumption? That shift might be a bad thing, but it's not an energy crisis.

I would argue that Triton is in fact higher-level than HIP. Plus, it is more specialised for specific use cases.

HIP is essentially the same as CUDA, CK is not a language but a library, and assembly is basically used in the Nvidia ecosystem as well, in the form of PTX.

There is absolutely nothing out of the ordinary here. Yes, it's multiple languages, but not any more or any different than what you'd use on an Nvidia platform (except obviously for the assembly part -- AMD's ISA is different from PTX, but that's to be expected).

While there's some truth to that, it makes a huge difference whether you have the choice between two (realistic) options or five.

> And it has always been thus. One votes for a representative

In the US, yes. In many (most?) other countries, many votes are for parties instead. This allows for better representation.

If that is indeed the case, then it was not a vote for what Republicans are doing now. It still points to a systemic problem: the system gave certain individuals the power to do things that they don't actually have a mandate for.

I'd say one of the major problems is that the voting system is built to de facto ensure a two-party system because it's a system in which individuals are voted for, with each race being a winner-takes-all for a single individual; as opposed to a system of representative voting. And a two-party system simply isn't able to represent the complexities of the world.

(This is of course in addition to other problems of the system which are also real.)