balsoft @ balsoft @lemmy.ml

Posts

4

Comments

311

Joined

1 yr. ago

For further context, opposing the war can and does lead to 10-15 years imprisonment in Russia, and Russian prisons are even worse than ones in the US.

The sad/scary part is that the missile research did indeed work pretty damn well.

What the hell is Trump admin trying to achieve here? China was already overtaking the US in scientific research. Randomly detaining researchers just because they're foreigners (among other things like defunding universities) ensures that in 5-10 years US will be hopelessly behind on everything from social sciences to medicine to physics, the list goes on really, and not just behind China but likely behind EU too. This really does seem like an empire in decay, however I'm willing to bet that people working on missiles to kill brown children will continue to be well-paid until the bitter end.

Anything more complicated than business logic in JS/Python sends LLMs into a guessing game that can take you those 3 hours to get out of. Try asking it to write embedded software in C, hardware-interfacing code in Rust, or any non-trivial TemplateHaskell.

That's not what I want though. I really enjoy jumping around the actual syntax tree of the code, e.g. "select the entire function body" or "select the next list element", stuff like this. It becomes the natural way of traversing the code after a short while. Also, Emacs is still single-threaded and thus quite laggy and slow at times; however I do like it a lot and have used it for a number of years (with evil-mode), before finally jumping to my own editor and then helix.

Nah. I was so annoyed by how primitive editors are that I started writing my own one, that would allow me to seamlessly traverse the AST of the code, rather than being stuck on the low abstraction levels of characters, words and paragraphs. After a bunch of misery making tree-sitter work with Haskell, and using it for a while, I stumbled upon Helix. It is pretty much my idea but faster and working well.

I guess it depends on the definition, but I'd say they absolutely are errors; if some function produces a result that is both unexpected for the user and outside the design criteria it should be considered an error.

I hate that I understood the joke immediately...

TBH this finally pushed me over the edge. It was surprisingly easy - just s/firefox/librewolf/ in my browser config file, and then privacy.clearOnShutdown_v2.cookiesAndStorage = false (I don't want to log in to all my work accounts every time I restart the browser). I didn't even notice the switch TBH.

We made a (so far internal) tool at work that takes your activity from Github, your calendar, and the issue tracker, feeds that to a local LLM, which spits out a report of what you have been doing for the week. It messes up sometimes, but speeds up the process of writing the report dramatically. This is one of those cases where an LLM actually fits.

I think it's best to get out of that cycle and force your body to wake up at the first alarm. Otherwise you're just wasting time on nothing - your brain doesn't rest properly in those 5-minute doses and you're not getting ready either. The way I'm doing it is to put my phone in a different room next to my bedroom so that I have to get out of bed to turn off the alarm. If you're managing to sleep through alarms it's probably not the solution for you, so maybe the QRAlarm recommendation made elsewhere in the thread is better.

They only dedup runtimes, not individual dependencies.

Not sure I follow you entirely, but I think we agree.

No, not quite. Flatpak is containers - it just stuffs every dependency that an application needs in a directory with no way to deduplicate or update independently. Gobo is a bit more nuanced, since dependencies are shared between applications when the versions match.

I think the main premise is that every version of every software has its own installation prefix. This allows you to mix&match different versions, perform atomic upgrades, etc. You can think of it as a proto-Nix. TBH I don't see much point in it now that Nix(OS) and Guix exist, or, if you don't like their purity, stal/IX.

The article is very light on details, but the numbers don't seem to check out at all. Back-of-the napkin math (assuming a square 1km × 1km solar array and total sun luminosity of 3.83e26 W):

    
1 km ^ 2 * (3.83e26 W) / (4 * π * (1 AU) ^ 2) * 1 year to TWh ≈ 11.94 TW·h

  

This is a "measly" 12 TWh of TOTAL energy delivered to the array over a year - not accounting for solar panel efficiency losses (20-24%) or the elephant in the room of transmitting this energy back to earth. For context, China alone consumed around 39 PWh (39000 TWh) of energy from fossil fuels just over the course of one year, 2023. The entire world consumed 55 PWh (55000 TWh) of oil energy in 2023 alone. It's not even comparable to the annual consumption of oil. If we consider the aforementioned factors, assuming 24% solar panel efficiency and an extremely generous 50% power transmission efficiency, we get:

    
1 km ^ 2 * (3.83e26 W) / (4 * π * (1 AU) ^ 2) * 24% * 50% ≈ 163.43 MW

  

Which is literally nothing on a national scale - it's less than a percent of the Three Gorges Dam output.

I think it puts some stress on the window, especially if you slam the door shut with too much force. I don't know how often this actually becomes a problem though, and in any case - you don't need an electronic door opener regardless.

The article says nothing about opening the doors from the inside; if you are ever in that situation, cybertrucks have a manual release under the rubber mat of the “map pocket” in the doors

I'm genuinely confused as to why you would prefer an electronic door opener. Is it just a gimmick for aesthetics? I've heard that it's to allow for frameless doors (in which it's better to open the window a bit before opening the door), but I've been in multiple cars that did that with a mechanical handle just fine, by simply adding a switch to that mechanical handle. It seems so stupid on so many levels: cost, reliability, repairability, and - chiefly - safety.

We can agree that China is a more stable partner than the US. My point is that party leadership can also change there, leading to policy changes - it just doesn't happen as often and there's usually no dramatic swings when it does. It's a scale, not a binary thing.

I think the more realistic argument is that CEOs have an inherent incentive to take big risks. If they get lucky and succeed, they get credited for the win, can put it in their portfolio and then demand better pay or move on to another company which will pay them more. If they fail, they can quietly resign, take the golden parachute and move on to the next company after a year or two as though nothing has happened. A big salary incentivizes them to keep their job, thus disincentivizing them from taking risks.