plqbfbv

I closed all my Reddit accounts after they instantly shadow-banned my 8+y main account after I shared a single gdrive link to reply to a user with actual data. I've been appreciated as contributor in various places for being helpful, thorough and precise, but Reddit doesn't care to have good content. It's simply not a good discussion platform (and the UI sucks).

Do you mean that the titlebar would be off screen so you couldn't move/close the window?

Not just the title bar, but often also e.g. half of the settings and the Cancel/Save/Ok buttons.

Basically the screen was 600 pixels in height, but the modal was designed after 2005 so it was 768 pixels tall, and you would get a cropped modal missing maybe 10% top and 10% bottom pixels that you couldn't interact with via mouse, and you couldn't resize nor move (either because it was fixed size, or because all corners that would allow you to do so were off-screen).

the OTAs break shit all the time

Uh? I can literally count the times my Model 3 2019 software broke something on one hand:

- when they redesigned the AC controls to make them more visually appealing but less functional (no button borders and no fill)

- when they decided to put air recirculation under auto-control and ignore the user's settings

- when they optimized the cellular connectivity and it took them a while to get back proper reconnect on loss of signal (garages etc)

- when they tuned sentry's sensitivity and there was some back-and-forth for a couple cycles between "record everything" and "record nothing"

None of this made the car undrivable or totally useless. I did hear of reports of early HW4 cars bricking their FSD computer, and Tesla replaced it.

In my opinion it's still a much better experience than the absolute guesswork of "what will my screen display today when I connect the phone? and where will I find Maps again?", based on software updates on the car AND the phone.

EDIT: also agreeing with the sibling comments: my 7 year old car got a lot of extra features since release, and most of them working very well at the first try.

Actual link speeds of Starlink terminals don't seem to be published in the open, but it's at least throttled to up to 310Mbps down/44Mbps up for Priority plans.

I had it last year in the Canary Islands (where I assume subscription is way below max capacity - no priority plan) before my current provider placed fiber, and I was getting up to 430Mbps down/100Mbps up with a consistent no-bufferbloat ~70ms ping, which is really not bad.

I guess in populated areas it can only reduce in speed as subscriptions increase, but in sparsely populated ones where fiber is not an option and 5G is crappy and priced similarly to Starlink, it's a valid alternative with no contenders for now. Not to mention that the ISP infrastructure it relies on does not depend on ground infrastructure, so if you're off-grid, you also don't depend on cabling/5G towers/etc for internet access.

I guess things keep getting more features and complexity but all of it is 80% done because it's "shippable" and updates are cheap.

This is most visibile for me on Windows/MacOS and complex web apps (e.g. GitHub and GitLab, including consoles of hyperscalers), where 80% of "normal" things work, then you need the last 20% and it's always not working as documented, only half-working, or just outright broken, and you need to find "temporary" workarounds that stay in place for years.

I feel this is being amplified by AI: tests come last (if at all) and are still written by LLMs, nobody really looks at them anymore, green pipeline checkmarks mean less than they did in the past.

Ordinary plebs trying to prevent climate change is like subtracting $100 from a billion dollars - it does not make any meaningful difference.

It stops making any difference if everyone stops trying. You compare this to subtracting $100 from a billion, but a billion dollars is composed of 10 million $100, and there's only a few rich people in those 10 million.

I think I saw a few years ago another version of this for IC and simpler CPUs.

For me the takeaway is:

- the libraries we use for chip design are small and reproducible, and we can reason about and account for their properties properly, so a designer can minimize design time by reusing small bits that are well understood and ship better architectures faster

- the optimal structures that would allow a design to minimize power use/maximize efficiency and at the same time guarantee highest performance are basically spread-out structures that humans could hardly conceive and reason about, because they look more like roots in a natural ecosystem rather than nice blocks of cut stone that you can stack (not to mention that developing a way to fabricate these chips can be harder)

Template libraries pushed chip design this far, and perhaps there some more to come, but we're hitting a wall where structures are becoming small enough that using libraries is starting to hinder the ability to scale further. These spread-out structures minimize interference and hot-spot creation by separating components that are normally "on" together, allowing better thermal distribution (and thus budget) and reducing interference by physical distance or "off" components, thus unlocking higher clocks and better efficiency.

What about the losses in the electricity transmission and the batteries and the conversion to motive force in the motor? Is it way less than that 40%? And wouldn’t there be more than 0% losses because refinery -> power plant shipping?

I once read this article and I find the diagram at the top conveys the point perfectly: https://insideevs.com/news/332584/efficiency-compared-batter...

EDIT: I once replied to the occasional "but that's optimistic for EV and pessimistic for ICE" poster: IIRC, making every EV step 10% worse (e.g. 15% energy losses in transport, storage and distribution) still has EV come out on top by a decent margin.

That chart also partially tells the reason why hydrogen is not going to be a solution for cars: you'd be swapping gasoline for a much greener but only slightly more efficient fuel, while carrying around bulky tanks and having to redo the entire transport network (oil pipes won't do). Maybe for things were work/mass works in favor (naval shipping would likely be one of those) and/or for storage it may make economic sense.

Technology is evolving, so perhaps some day we'll be able to produce and store hydrogen with losses similar to the EV chain, but I doubt it: hydrogen is the simplest gas of all and it's extremely reactive, plus its molecules are tiny (it can leak through metals), so tanks and pipes need to be done with different materials, and I doubt we'll find better containment materials given the physics.

I have the Fairphone 5, for which they picked the IoT equivalent of the mobile chipset so they could ensure 10 years of driver availability from the vendor (8 years of majors + 2 of security patching IIRC).

Fairphone 2 came out in December 2015 and saw the last release in March 2023. That's 8 years of software updates when Android devices from that era barely got 3, and you had to pick the flagship.

Kernel updates used to be bound to Android versions because of how the kernel modules development was handled, not really limited by the company or the hardware. I owned a OnePlus 5 and the custom ROM was stuck with older kernels for a while too, until the community stepped up to port it, because there just wasn't an easy way to build kernels with updated custom vendor modules for the hardware. Google addressed at least that part thanks to the Kernel Module Interface (read e.g. https://arstechnica.com/gadgets/2021/09/android-to-take-an-u... ). So newer kernels may still lag a bit based on the manpower available and priorities, but they should be easier to do.

Now that replacement cycles have become longer because hardware is good enough that you don't need to change phone every year, and Android supports KMI and other features that make maintenance easier, more vendors decided to extend support, so at least there's some more choice.

From the environment point of view, I can now finally easily repair my phone. A friend of mine always had bad luck with pixels, so bad that he went through 4 phones in ~5 years (2 bought, 2 replacements). I am at my third phone in 12 years (OP1, OP5, Fairphone).

There's no way to verify the integrity of the system, and any malicious app can just grab your banking credentials or enable criminals to unlock and drive away with your car.

I get that Google doesn't want to be sued for failing to protect its users and indirect users of the mobile phones sold by other companies, but for advanced users there should be an option to update the signing keys used by the bootloader, so that you can unlock, flash your custom ROM, update keys, and relock bootloader. Such a phone should still be considered "trusted" by Google Integrity APIs. But currently there's no way to do this, so basically you don't really own your hardware.

I gave up on custom ROMs trying to extend my devices' lives and bought a Fairphone instead, so I have the assurance from the vendor that I will have software updates for a very long time.

Sure people care about the environment, but it is always lower than all the other factors in their personal list of things they worry about.

Emissions scale with performance, and inversely of fuel efficiency. So the environment may not be the most important point, but I'm pretty sure fuel efficiency is high on the list when you're picking a compact or long-range car that is supposed to be fuel efficient.

Also, by advertising as compliant to green specs something that wasn't, means people may have been swayed to purchase irregular cars despite them not being really green, only due to the fact that they may have received rebates and contributions for the purchase, regardless of whether "being greener" ranked high on their decision metrics.

Early adopter and subscriber (I think I saw the link here on HN), have been happily paying for 4 years now.

It does one job and does it well, and you have additional features if you want/need to configure them or tweak the results to your liking.

AI is an add-on that you can elect to use or not, and I enjoy just adding a "?" at the end and getting a proper AI response instead of an hallucinated AI summary by default that covers the entire page.

2. Firefox's tab crash recovery isn't as solid. I use chrome with fully persistent tabs, and its a gigantic pain if I can't re-open them

I normally have 5-50 tabs open (so perhaps on the lower end), but I can't recall the last time I crashed a tab in the last 3 years. I also use persistent/pinned tabs and never noticed issues.

Solar is a massive maintenance liability that a majority of buyers will avoid

My parents put up solar in 2005 when it was very expensive. The original panels, inverters, cables are all still there, nothing had to be replaced in the last 21 years. The only maintenance task is a recommended (not mandatory) yearly cleaning that is probably in the 300€ range and takes maybe 1h, but since 5 years they have 2x EVs and 2x Powerwalls and are self-sufficient for 80% of the year, so that cost is negligible. The area where they live also has enough rain that if they skip cleaning for 1 or 2 years, the average total output is still close to the max they can get.

Fair, I had a quick look, 600-700 is likely for the big cities areas.

Moved out last year after 10y in the Zurich area.

There's always been a pull-and-push between getting skilled workers and protecting the internal labor market. Right-wing political parties never made a secret of the fact that they hated immigrants, because they stole jobs and redirected/exported money that would have otherwise been received by Swiss. IIRC this was historically mostly felt in Ticino (the southern region), where Swiss companies sourced very cheap Italian labor by undercutting Swiss salaries by a lot, shrinking the job market for Swiss people (a Swiss can barely get by in Switzerland with an equivalent Italian salary).

Switzerland is geographically in the middle of Europe, but it's not part of the EU. This allowed the country to thrive outside some of the more restrictive EU regulations and keep its own currency, but because it has a smaller job market that can barely replenish the high-skilled workers pool and is often in defect (not just finance bros, but also doctors, for instance), it always had to import workforce from neighboring countries to some extent. Over the last 40 years Switzerland basically opened up to more-or-less follow many EU rules and put in place agreements to have a play in the same market and be allowed to easily keep importing people it needs.

This initiative as I understand it would be equivalent to a Brexit (because sooner or later the cap would be hit, considering more housing keeps being built), which would undo 40 years of openings and IMO greatly weaken the integration with EU, and as a result the country as a whole.

Yeah, but while Belgium is basically a huge plain, Switzerland is 60% Alps.

If you account for that, the effective density of Switzerland on the usable area is 600–700 people/km².

I had to start using noTunes years ago because whenever I accidentally touched or removed my (Sony / non-Apple) bluetooth headphones, iTunes/Music would pop up without fail. Apparently they sent a "play" button press, so the application would pop up. Because it's a system application, you cannot delete it, there's no way to override the key, there's no way to disable it. So noTunes has been running for 7 years straight on my Mac.

why companies like Tesla deliberately make their cars inconvenient

My 2019 Model 3 has been the most convenient car I've driven by far. Other companies saw Tesla get away with it and tried to replicate it, but given they had basically zero software UX experience and vision (and still don't, on both), the result ranges from kinda usable to absolutely terrible, with most trending towards the latter.

And I totally agree, but I still use mirrors because when I need to pass and park next to a pillar (my garage has a few) I can never tell how far I actually am from it through the cameras, whereas the mirror will tell me without lag and with absolute precision whether I'm about to scrape my car or not. Not to mention that it's very difficult to tell depth from cameras, which is what I need the most when the back axle has passed the pillar and I should start turning in very tight spaces.

You can definitely do it with a Tesla. Disable Auto, select the airflow direction and vents, select the fan speed you want (1-10), select the temp you want. Fan speed will not change with temp if manually selected.

It's really hard to justify $1000 for it when the Switch 2 is $450 (soon $500).

I had a Nintendo Switch 1 and sold it. My wife wanted to play the latest Pokemon release, she did for 2 months, then it sat unused for 3 years in its charging deck.

If you like some titles that Nintendo has to offer, then sure (I never understood the appeal of the Zelda saga for instance).

If you want a portable PC optimized for gaming with a huge library and where you can also do anything else you want, including modding it, 1000$ is still a bargain for what it offers. EDIT: and the library will keep expanding as Proton keeps maturing.

It does take a while for all warrants to be in place, run the investigation and tie up all loose ends.. 6 months is fast by bureaucracy standards.

The fact that he happened to be in NY is a lucky shot for the prosecutors.

Polymarket accounts should be untraceable

The funding/retrieval chain normally isn't if you want to spend the money in real world goods and services.

if an employee can predict specific search results to such a granular extent

It's not a prediction, it's a query in a tool/database that's supposed to hold counts.

There's no rigging here - it's just a summary of historical data being accessed and used before being public, in violation of confidentiality agreements. If he had just sourced the data without acting on it, the system would have still logged somewhere, but likely nobody would have looked at it. My understanding is that the last couple weeks before being released won't dramatically change counts unless an extraordinary event happens that drowns the last 11 months of data.

What prevents car manufacturers from taking a normal-looking body style and electrifying it? Seems popular with the aftermarket mods. Every hybrid and EV I can think of looks like a suppository.

The fact that even though an ICE engine is only 35% efficient, the energy density of fuel is still much higher than gas. Gasoline is 50x more energy-dense than an average EV lithium battery (taking a Tesla Model 3 as example). That's why you can fuel 9L of gasoline or charge and carry around ~500kg of batteries for the same range.

So in order to make the car efficient and travel 500km you have to bend the shape to be as aerodynamic as possible. That's where retractile/flush handles also come from, it wasn't just a gimmick or wow-factor on the Model S when it launched. It's still like that years later on all models because it still makes the car more efficient.

The research and push for higher densities is finally starting to pay off, so in time less efficient but more appealing/classic designs may start to emerge again, because you're less constrained by how much energy you can carry per same weight/volume.

Just like parent poster, there's a few things I pay for, and those are Kagi and Fastmail :)

Fastmail is... FAST. The page refresh is basically instant, and the UI is super-snappy, well refined and something like 2026, while Gmail hasn't changed in... 15 years? It supports a lot of features (identities, masked addresses, ...), not to mention that I can have my domains properly hosted on it, for 5eur/mo.

such that they can use the right set of sensors in the right environmental conditions

Because this part is really hard, and that's why Tesla abandoned the fusion approach. You cannot possibly foresee all the conditions in which LIDAR or any active sensor will malfunction/return wrong data/return data that's only slightly off for that ONE specific time. And even if it doesn't, you need to trust it to not return noise. And when it does return noise, how do you classify it as noise?

Cameras are passive sensors - they get whatever light comes in and turn it into an image. Camera is capturing shapes that make sense to the neural nets: it's working. See all black/white/red/cannot see any shapes? Camera is not working, exclude it from the currently used set of sensors or weigh it less when applying decisions, because it's returning no signal (and yes, neural nets have their own set of problems).

EDIT: cameras also provide more continuous context: if 1 pixel is off, is clearly bright red in a mostly-green scene where no poles can be identified, the neural net will average it out and discard it as noise. If 1 pixel says "object" in LIDAR, do you trust it to be correct? Perhaps the ray just hit a bird or a fly, but you only see a point, it's a lossy summary of the information you need.

I would NOT be using Starlink for remote vehicle teleoperation even as a fall back.

I had to use Starlink last year, and latency was way more acceptable than expected even when under load (I did try to analyze and remove bufferbloat). Considering Tesla could likely get priority bandwidth from SpaceX basically for free, that would mean good latencies (I had 90ms tops in speedtests). Anyway you tell the car where to go, but it's the car following the path you draw for it and following traffic rules and collision avoidance, you're not directly driving the car. Even 1 second latency with 2s round-trip would likely not be a problem in these conditions.

Why do I need a cluster?

I run a single-node K8s cluster on a dedicated server because it's way cleaner to manage than the previous mess and mix of docker compose + traefik routing + random stuff installed as package on the host.

I can create "vhosts" for practically anything in a declarative manner, and if the cluster blows up, I have 5 small scripts to bootstrap it and all I need is `kubectl apply -k .`.