nzentzis

While I can’t speak for GP, I had a similar reaction - though admittedly I wouldn’t describe it as “relieved,” more of an “okay, good to know this isn’t anywhere near the top of the priority list” feeling.

Essentially, in comparison to the other potential/likely effects of climate change I’m aware of (mass deaths of pollinators causing a collapse of the global food infrastructure, large percentages of the world’s arable land becoming non-viable leading to mass famine, heat waves bringing lethal wet-bulb temperatures to large populated areas, collapse of the AMOC, increased wars and global conflict due to space pressure, large-scale droughts and water scarcity, etc) it just doesn’t seem that bad. It’s awful and terrifying, to be clear, but it doesn’t really compare to some of the other effects we’re going to be dealing with over the same timeframe.

I'm using a janky CUDA miner that's getting around 6GH/s on a 3090. My code is kind of bad though since I hacked it together in a couple of hours last night, so I wouldn't be surprised if a proper/competent implementation got 2-3x the hashrate mine does.

Getting into the top ten is still not too hard at this point, if you wanted to. I hacked together a super basic parallel CPU miner quickly before diving into CUDA, and the CPU one still gets a value that beats #7 within a few minutes.

Assuming hashes are random (which is a reasonable assumption, considering it's sha256), every extra leading zero cuts the probability of success in half, meaning double the candidates searched to find a matching value. So each additional leading hex 0 would take on average 16x longer than the last.

There’s not really any viable options other than Nvidia for GPU compute, despite AMD’s efforts on ROCm. If you want to do compute tasks, you need an Nvidia card with Nvidia drivers, and Wayland won’t support your usage.

The problem is that there aren’t enough digits of precision to represent the answer. The exact answer would be 999999.0000001, but let’s say you only have space to store six significant digits. The exact answer is way more precise, so you need to round it to the nearest representable value, 999999.

Floating-point rounding issues are the same thing but with binary digits instead of decimal ones.

The “oh, just don’t buy a GPU from NVidia” point is one I see made a lot when discussing Wayland’s failures, but it completely misses the fact that CUDA is currently the only viable option for GPU compute in many situations. If any significant part of your desktop usage involves such compute, that’s not a solution - or rather it’s a solution that renders your system less capable than it was before.

Though in a sense that’s consistent with Wayland’s general “you’re holding it wrong” approach of shifting blame for any problems onto the person reporting them and concluding that anything that doesn’t work well isn’t a valid use-case anyway.

I want to be able to justify using Google Cloud for projects, but the fact that it’s a Google product is the main obstacle to that. If history is anything to go by, relying on Google products and services seems like a great way to set yourself up for major problems down the line. Either the product you’re using is shut down on short notice or some automation flags a false-positive and you’re left screaming into the void trying to get yourself unbanned.

Thank you for this recommendation - I'm planning to switch platforms in a few weeks, and blocking YT ads was one of the big open questions I still had.

On the contrary, there's no reason to believe the exorbitant initial cost would change at all. If anything, this would make it worse.

The reason textbooks can cost as much as they do is that demand is very inelastic - students don't have any real alternatives to purchasing textbooks (in most cases.) As a result, there's little downward pressure on textbook prices.

Without any reason to lower prices, publishers would absolutely take this opportunity to maximize profits by raising secondhand prices and taking most or all of the difference. Students will be forced to buy them anyway, so why wouldn't they?

Can you recommend a good reference/source for translating legalese? I'm not sure I've ever found a good source for determining which words should be interpreted normally and which are landmines.

The ambiguity about what words mean may be part of why people assume patents are so broad. Most people know that some words and phrases, when used in a legal context, have vastly different (more specific, broader, or even completely disconnected) meanings than what you'd expect in normal writing or speech. Without knowing what those are, the safe approach is to ascribe the least favorable possible meaning to every word. "Does language X mean Y" becomes "could language X possibly be interpreted by someone who doesn't understand the material to kinda vaguely reference Y," and you get the type of broad assumptions you're lamenting here.

Should be F3, I think (works on my system and I don't think I changed the defaults).

If that's true, and consensus is only starting to come together now, how is the Wayland ecosystem considered ready for mainstream usage?

From the perspective of someone happily using X11 at the moment, Wayland (or whatever your preferred term for "the loose association of compositors, protocols, extensions, and nonstandard hacks making up the Wayland ecosystem" is) looks like a failed attempt at building an ecosystem with proponents who are now trying to push it on everyone else in an effort to get the rest of the open-source community to solve the problems they created.

Every compositor is doing their own thing, application and framework developers need to implement basic functionality in one of several different ways depending on which DEs/compositors/WMs they want to support, some stuff has no replacement at all, and we're going to have to throw out the entire X11 world in exchange for... smooth DPI scaling and vsync? Really?

I honestly want to switch to Wayland - some of the stuff I've read about the X11 codebase is terrifying - but the cost of doing that, throwing out the entire desktop world, and giving up legitimate use-cases as "you shouldn't want to do that" is just too high, and the benefits are minimal. I'd honestly be happy to switch, but the whole ecosystem feels like it's a decade or two from being ready to go.

A lot of the hate Wayland gets stems, in my view, from the way it's been pushed on people. Users who aren't invested in the ecosystem and just see people pressuring them to switch to a loose collection of half-finished software that doesn't properly replace what they already have.

First off, I hadn't run into Hammerspoon before, so thanks for that - it looks like it might help with some things.

Most things I've tried and failed to do involve silent/background interaction with apps. For example, my company's VPN software doesn't handle sleep well at all, so I wanted to automatically disconnect before sleeping. I couldn't work out how to do it. The Mac Automation stuff (Automator and AppleScript) work if the app supports them, but my experience has been that support isn't necessarily reliable. Electron apps, at least, ignored my attempts last time I tried.

I figured out a way to do things using keyboard/mouse automation, but that fails in any context where I'm either actively doing something or where input isn't being accepted (e.g. when the lid is closed or when I'm actively editing code). On Linux I can at generally resort to Stupid X11 Hackery to make that kind of thing happen, though it sometimes needs WM integration if the app isn't cooperating.

Seconded. I use both on a regular basis - a Mac laptop for work since I have to interact with the MS Office ecosystem there, and a Linux desktop/workstation for gaming and personal projects.

Between the two, I prefer the Linux machine on pretty much every axis. MacOS feels far too tied to the GUI-oriented way of doing things; it's often difficult or impossible to automate tasks/routines, and troubleshooting when things go wrong is painful.

I'll second the "Valve's been amazing" comment. The majority of games I play are native, and all but a few of the rest work fine under Proton. Compared to when I started using Linux on the desktop a decade ago, the gaming situation is night and day.

Regarding Fallout 4, though, I played through it using Proton a while ago. As far as I can tell, the mouse thing is just an issue with the game - I found multiple sets of instructions to fix it that assumed you were using Windows.

I think C and C++ have enough problems with accidental undefined behavior already without making aliased pointers into UB.

> Look at the lengths involved in getting A* to run on Curiosity

Can you link to something that goes into detail? Googling it doesn't turn up anything relevant, but it sounds like it'd be interesting to read about.

Both Windows and Linux (via PulseAudio) provide this. I'm not as familiar with the Windows side of things, but on Linux I can also redirect each application's sound to different audio devices.

I just tested this on my Linux workstation.

2000 threads does nothing - everything's still responsive, and the process is shown as using 0% of the CPU.

16,000 threads uses ~30% of a core, with a ~136MB RSS. The system still handles it fine, though, and everything stays responsive.

At 20,000 the program panics when spawning threads, with the message "failed to set up alternative stack guard page" due to EWOULDBLOCK. I'm not sure exactly what limit it's hitting, though.

At least for the purposes of comparing to web apps, the criteria would be: responsive/low-latency (responds to input quickly), fast (completes tasks quickly), uses resources proportional to the functionality it provides, and doesn't often hang or spend noticeable amounts of time waiting for a network request before responding to an action.

Applications that meet these criteria include: Thunderbird, KiCAD, VLC, Vim, tmux, Blender, evince, Handbrake, and Pidgin.

I keep hearing the "web technologies are fine, you're just using them wrong" refrain over and over again, but it's ultimately not very convincing. If it's true, then where are all the good web applications? I've certainly never seen any of them. Can you give an example?

If nobody can get Web technologies to deliver a good experience, then I'm not sure it matters that much whether or not it's possible in theory. Delivering a theoretically great user experience won't get you anywhere unless it also translates to practice.

My main problem with it is limited software support. Xmonad isn't available and as far as I can tell what support exists for screen recording and screenshots is half-baked at best. I haven't seen anywhere near enough problems with X11 to make switching window managers worth it, and the screen recording thing would be a massive pain to work around.

This doesn't apply if you want to run CUDA-dependent software. I've generally gone for Nvidia for my personal machine since Torch has behaved oddly on AMD cards in the past.

It's true that Nvidia doesn't support Wayland properly, but that's not really an issue in my opinion. Wayland still has its own problems that mean switching from X11 isn't viable yet.

I think I partially agree, but that it mainly depends on the audience. I've found that parsing and recursion appear in more areas than just compilers, to the point where it'd be relatively difficult to avoid them. Machine code generation and the underlying implementation of high-level primitives (closures, tail-call recursion, and the like) hasn't, at least in my experience, been "naturally-occurring" to the same extent.

In terms of creating a "learn compilers from scratch" resource, Crenshaw's approach is definitely better. The trade-off would be that it'd take longer to get past the "writing a recursive descent/LR parser" phase, and it might never get to higher-level language features at all depending on the input language you go with.

Currently I'm using a modified version of the after-dark[0] theme for Zola.

[0] https://www.getzola.org/themes/after-dark/

Generating 64-bit code would be simpler in many ways, but I decided not to go into it until the basic language features are put in place. The extent of the changes involved in switching to amd64 will help directly show why having an intermediate representation for the generated code would be valuable.

Besides, if you're looking for a high-performance Scheme that fully utilizes all available system resources, there are definitely better options available. :)