milch

I do think there's a point where introducing different failure modes is harmful. Say they end up causing more pedestrian crashes and more head-on & angle crashes (such as the one in this article), but less loss-of-control and fixed-object crashes. That would make it safer for occupants but less safe for non-occupants of the self-driving vehicle. Since the occupant is the ones engaging self-driving mode they should be the one carrying the risk, so IMO even if it reduces number of deaths overall it should not be allowed

I have an LG OLED as well, and to be honest I don't care. It is pretty obvious that every TV manufacturer does it by the fact that they all have to make their TVs "smart", perhaps save for the tiny/unknown ones that don't have an ad department or a contract with an ad company. Five minutes of using any smart TV interface should be able to dispel the notion that it somehow exists for the benefit of the user, and not for other reasons. My TV simply never goes on the network, which makes this a non-issue

Isn't the whole point of UBI that it is universal and equal, e.g. everyone gets XX$ no matter what? If it is tied somehow to behavior then it is not UBI anymore, but more like a tax credit

Maybe I don't remember the names, dates, and others correctly but I don't think they had things like

> Wi-Fi and Ethernet clients are equivalent peers on the LAN. > Wi-Fi and Ethernet clients aren’t equivalent peers on the LAN.

To be honest I don't know what this is supposed to be telling me. They can or can't be on the same VLAN? They could have different or the same QoL? If we hammer traffic on both at the same time bad things may or may not happen?

> Only IPv4 is running on my LAN. > Only IP is running on my LAN. > Okay, only IP, IPX/SPX, Phase IV, AppleTalk, VINES, NetBEUI, or SNA are running on my LAN. > Seriously?

Practically I'm not sure how this affects someone writing a networked application. Sure, IPv4/IPv6, but if I get a bug report in the year 2026 that my application is broken because it doesn't work on the AppleTalk network, it gets a WONTFIX.

Sure, maybe if I'm writing a low-level network driver or maintaining a mainframe application these are things I have to care about. I think that may be the difference between the other posts and this post - non-experts in Onomastics and Chronology deal with names and dates all the time, hence resulting in issues like government websites not accepting Ng as a last name because it is less than 3 characters or not allowing entering February 29th as a birthday.

As someone who has done OSS work maintaining a project of similar size, and who works with people maintaining projects 10x larger, IMO this is totally fine. Maintainers deserve to get the bag. 99.999% of users won't pay for it, even if their company's product depends on it or if every dev in the company uses it. At the end of the day the people making the product have to get paid somehow, and if people don't want to look at sponsorship posts or ads they can go to something like https://www.git-tower.com/ and just pay for the product directly

I've had my code editor's configuration set to auto save for at least 15 years at this point. I've never had the problem of not being able to throw something away that I wanted to throw away - either it's in the git history (broad edits), or it's in the undo-history (fine edits). It helps to have an undo history that survives restarts and that preserves branches, which I guess may not be a default for every editor

We don't have carpet, and I use the grounding armband from my iFixit set. I've never had an SSD die either even when swapping them between various systems over the years. My oldest M.2 is about 10 years old, a 120GB that I now also use as a boot & VM drive for my NAS. Maybe I'm just unlucky ¯\_(ツ)_/¯

Though if a component dies due to ESD from an imperceptible spark then the manufacturer skimped on ESD testing

The cheapest compatible one I can see new (and one of the only ones...) is currently at 120: https://www.newegg.com/p/pl?N=100007625%20601413455%20601409...

That's actually not bad, last time I checked the only ones that were in stock cost 300+ (IIRC all x570). My plan then was to maybe get a new case and rebuild it on a micro ATX board which are generally more available and seem to have a few decent options in the ~80 price range.

Anyway, again it's a question of is it worth it spending several hundred and the time it takes to rebuild. It currently "works" fine except it takes 5 minutes to boot and is unable to save any BIOS settings like XMP, fan curves, and so on (and yes I changed the CMOS battery - there's a long thread on the ASRock forums of others having the same issue even after the battery is swapped). The non-optimal settings probably eat a few % of performance but other than that it's mostly annoyance, so it still feels like if I'm going to invest any more in it, it better be more worth it

I bought from Newegg both times. The first time it was an ASRock MB that was dead, and to be fair, the issue here was that some part was on backorder so by the time everything arrived and I put it together I had to deal with ASRock RMA rather than the Newegg returns department. Typically I choose pretty high end components as well, I do Corsair PSUs & RAM, Samsung SSD, ASUS GPU. Definitely learned my lesson on the MB and went with ASUS for the latest build as well... though that MB's DIMM_B slot was dead. I'm just glad I ended up getting 2x16G of RAM right away instead of starting with 1x16 like I originally planned, because I would have never realized the problem until getting the second stick

Yeah, just barely, and only if you don't value warranty or build time. I'm also not super sure you could get something together that is somewhat compact like the steam machine and still be cheaper.

I've built two PCs in the last 5 years and both have had parts that were broken from the factory... There is something to be said about being able to hand off troubleshooting and replacement. I had ASRock pointing fingers at everything else to the point where I had swapped out CPU, PSU, and RAM, until they relented and replaced the motherboard (the replacement which is broken again now 5 years later by the way...)

I've been in the same boat since it seems something about the memory controller on my gaming PC's MB is broken, but price wise it feels like it just doesn't make sense. I could get something from eBay relatively cheap but then I'd probably spend an hour disassembling and reassembling just to maybe learn that the eBay unit has some equivalent issue or is slightly broken in some other way. I could old new stock for outrageous prices. Right now I'd rather just wait until either an upgrade would be REALLY worth it or at least prices come down.

> As far as I can tell, non-stick is just a ploy to get people to keep buying new pans as it wears off.

This is one of the biggest reasons my wife and I switched away from non-stick too. We moved from replacing non-stick pans after a year of use to our carbon steel pans working better after 5 years of use than when we got them... or buying 80 year old cast irons from a thrift store for 5 bucks. We have more types of cookware now, but none of them have an expiration date

My wife and I started doing this about 5 years ago or so. Since then, we gradually acquired more carbon steel, stainless steel, cast iron, and enameled cookware. Initially we were reaching for the non-stick for all the typical reasons, e.g. making eggs, but over time we found ourselves reaching for the non-stick less and less. It turns out that the "seasoning" on the cast iron and carbon steel improves their non-stick properties over time, which at this point is good-enough for anything... I can cook eggs, pancakes, a piece of salmon without breaking the delicate skin or shredding the fish, and so on. As I see it, non-stick pans have two benefits: they require less maintenance (you can stack them in the sink and deal with them whenever, which you shouldn't with uncoated metal pans after cooking anything liquid/acidic) and you can cook more easily without using any fat. For us, neither matter much, so about 2 years or so ago when our last non-stick pan needed to be replaced we decided not to get a new one and just get rid of it.

With the Chinese models now being 3T too I don't think they have to. Speed/cost wise this already doesn't seem like a viable model. The biggest threat for the economy IMO is if they challenge the whole Nvidia/Datacenter infrastructure money side, which this pretty much doesn't. More likely they will ban running LLMs locally in some way

And somehow just because the store says it's 48gpbs doesn't mean that's true either... I have a cable that is explicitly marked as 48gbps, 8k60, 4k120 compatible and yet it frequently ends up dropping the connection. Swapped to a different 48gbps cable and it works fine. Funny enough when I display the negotiated FRL speed it's only at like 27gbps, so should have plenty of headroom.

At some point my older Mac charger that I was using for traveling died, so as a replacement I got a few of these styles of adapters: https://www.amazon.com/dp/B08YS7QHM9 (not the exact one I have - but you get the idea). I pair it with one of these: https://www.anker.com/products/a2687-anker-prime-charger-160... which is tiny for the amount of power it can output, which has been enough to cover both me and my wife while we travel. Many US plug chargers allow their pins to fold in, including this one, which makes for a fantastically compact setup when paired with that style of adapter.

Depending on where we are going and what we are doing, I also bring this thing: https://www.amazon.com/dp/B08YDPB8RZ which is kind of an extension cord + USB charger + all-in-one converter as well. The plug is at a right angle and very flat, so it can fit behind furniture and such that straight cables can't.

I built myself a new PC recently and jumped straight to installing Linux on it. Unfortunately the experience isn't quite as plug and play there either... In terms of everything I had to do:

- Disable iGPU in BIOS because on a cold boot the system would "latch on" to the iGPU (iGPU was GPU 0 and dGPU was GPU 1, which some programs I guess interpret as "GPU 0 == primary GPU"). Everything is fine on a warm reboot where it assigns dGPU to the 0 slot and iGPU to the 1 slot

- Get a special dongle to get 4k120Hz support over HDMI because the driver doesn't support it. The dongle should also enable VRR but needs multiple kernel patches to do that. Applying the patches causes varying levels of graphical glitches

- On waking from S3 the HDMI audio is dead. I tried various kernel arguments related to power management that various people had reported working for them but ultimately none of them ended up working for me so this one is still broken. Gave up on HDMI audio out and used an SPDIF cable instead (my AVR has been a magical problem solver here but most people aren't going to have a separate device to do that for them)

- On waking from S3 the wifi doesn't automatically reconnect. Solved with a systemd unit that just toggles it off and back on

- The audio out kept changing away from the one I set. As far as I can tell there's no easy way to set the "default" like you can in macOS or Windows, so I had to fudge the priorities of the built in devices a bit with a Wireplumber config

While debugging all the issues I've been bouncing back and forth between various kernel versions and patches. The fact that you can do that is awesome, but the fact that you have to do it to fix and debug issues is unfortunate. Just as an example, while debugging the audio problems I found a ticket that looked like it could be related on the amdgpu driver issue tracker where someone reported a similar problem, and the response from AMD was "please bisect the kernel"... That's just not something that most people are going to be willing or able to do. AMD GPUs are even supposed to be the well supported platform compared to Nvidia. I will say though, this whole debacle with the dongle sure had me wishing I could just stick an opaque-but-vendor-supported blob in like you do with Nvidia and get HDMI 2.1 working out of the box.

At least on Windows, IME Microsoft ironed out a lot of these types of driver issues in the Windows 7/10 era so that things like sleep just tended to work. It's been a while since you had to track down the right driver for your hardware manually. They just kept going downhill from that point forward with boneheaded decisions like sticking React Native or whatever in the start menu and forcing shovelware down people's throats. Even though I was having all these problems with Linux I personally won't go back to Windows, but I'll think twice about recommending it to my non-tech friends and family.

I think there's a world of a difference between "you wrote some code and it sucks" to "you wrote some code and it sucks because you have beginner energy and live in a fever dream". The former can be an objective statement without attribution to personal failings, while the latter... not really. If I put the latter as a PR comment to my intern in the morning I would expect a meeting invite from HR by the afternoon at the latest

How historical? Compilers that have translated from a source language into C and have left the C-to-bytecode translation to another compiler have been around for a long time, as have compilers that translated from a source language to Assembly

Very interesting article. Would've been fun to see the comparison between native interpreter & JIT-on-WASM on iOS as well

I'm never mean but sometimes when Claude does something especially boneheaded I just hit it with a single "bruh". That usually triggers an automatic "You're absolutely right -- I shouldn't have X and followed your directions more closely, let me revert and do Y instead"

Looking at their blog about it, I don't think Evolver has any claim to be honest. They make such generic claims as that session history was copied because they implemented their history in some `.jsonl` files while Hermes has a session history using sqlite FTS. Evolver better have some attribution crediting Anthropic for their session history feature, because as far as I can tell they've had that implemented since early 2025! Using .jsonl files no less, coincidentally the same way that Evolver decided to implement session history.

I was expecting them to show code samples or something, but to me it looks more like these were two projects that ended up coming up with similar looking features, even with different implementations of these features. I don't see how Nous Research owes them any explanation, especially since at least from that linked issue they were coming in really hot with accusations of plagiarism. If I owned a large OSS project that amassed issues/discussions/PRs at a rate of ~4k per month I would probably also treat this kind of thing as spam.

The limbs you can deal with, at least. I've done century rides in 100°F+ weather wearing UPF 50 sleeves the whole way... it's actually quite nice because you dip them in water, and they end up staying saturated for a long time. Together with the airflow from the ride it keeps you cool as well as protecting from the sun

IIRC Anthropic models haven't all been available on day 1, so it does feel like a deliberate choice, especially since they are partners/part-owners like you say. No one would have bat an eye about some corporate PR thing "blah blah mythos is too powerful and too intelligent and so we've decided to focus our capacity on Opus 4.6/4.7/4.8 for now until we have the proper safeguards in place blah blah"

Why would they have launched Fable on Bedrock if they knew they were going to be shutting it down a day later?

Instead NHTSA still allows automatic high beams, which blast you instead in the US... Or just regular old drivers that don't realize/care. Even as a pedestrian, I constantly see people driving through my neighborhood with high beams on... and these are side streets with a 20mph/30kph speed limit. Anecdotally at least people in Europe are a lot better about only turning high beams on outside of city limits

Look at the very first point on the archived vote leave page: https://web.archive.org/web/20160620214900/http://www.votele...

Sure, you could argue that they didn't mean it would be positive for the economy to save that money, but "we will save 350M/week" is what's on the buses and their website. Even if we assume the average voter clicks through here and reads everything point by point, or goes onto the website in the first place rather than by the headline, it is at the very least heavily implied... Otherwise what is the argument?

Weirdly I find myself writing more code lately, not less. A lot of it is AI assisted but a lot of it also isn't. I find that where previously the management chain wanted things meticulously planned out, and we had to have 17 meetings with different stakeholders and principal engineers to approve the system design.

What hardware do you need to run MiniMax M2.7 230B locally?

In fact when my company awards me stock the government takes half anyway!