funny a company is expanding production instead of sitting around waiting for the chinese to eat the market

jacs said:

That’s a lot of mummy movies… though it’s been almost 100 years since that first one (assuming that was the 1932 movie) so it’s less than one a year.

Click to expand...

The oldest one I have is titled, "The Eyes of the Mummy" from 1918.

The headline makes this sound more sweeping than the evidence actually shows.

Researchers initially reviewed 1,371 responses but analyzed 959, leaving 412—30.1%—outside the final adult analysis. That does not automatically invalidate the study, but the reasons for those exclusions and how respondents were recruited matter when interpreting the results.

These are relatively small absolute numbers—268 adults and 63 caregivers reporting the behavior—not a direct count of millions of Americans. To be fair, a survey of roughly 1,000 people is not automatically weak. But sample size alone does not establish that an online sample represents the country, and reporting on the research acknowledges that participants may not represent the entire US population

And the medical system deserves scrutiny, too. Researchers found that people were more likely to overstate symptoms when clinicians had previously dismissed or downplayed their concerns.

The fair conclusion is that this survey identifies a concerning behavior worth studying—not that it establishes how often Americans lie during medical visits, or justifies putting all the blame on patients while overlooking failures in access, communication, and prescribing.

Another day over 2000

They had fun with Halloween!

Login to view embedded media View: https://bsky.app/profile/volberg.bsky.social/post/3mz7govzff22j

w00key said:

Since Opus 5.5, users on Claude subscription are getting pretty much all they can use unless they go tokenmaxxing and doing crazy large projects, stuffing work that would have taken whole teams weeks to prep in a single weekly quota.

Click to expand...

Yeah, I’m there… but… now that I’m no longer blowing through my quota doing the kinds of things I did before, I’m slowly starting to rethink the volume of work I’m giving it. Things I would have held off on or down-scoped I’m more likely to tackle head on now.

Broadly speaking it takes time to figure out how to most effectively use a new resource. And getting 20X the amount of tasks out of the same quota is qualitatively a new resource. Over the next few months I think the narrative will increasingly be people coming to terms with the fact that they can give AI more ambitious projects than they did before and not run out of tokens.

Da Xiang said:

I count 64 Mummy movies that I can identify.......

Click to expand...

That’s a lot of mummy movies… though it’s been almost 100 years since that first one (assuming that was the 1932 movie) so it’s less than one a year.

Drizzt321 said:

You mean Node2Shell (CVE-2026-111497)?

Click to expand...

No idea, actually. It was more the "should we be doing this on bare metal with access to All The Things?"

Ban me, will ya?

TheOldChevy said:

AI replaced Reddit for a big part of what it was worth when developing. For other subjects where human remains valuable, we have other solutions.

Click to expand...

You mean AI answers stolen from humans on reddit replaced reddit for you.

Stromlo said:

Help me out here, I (sometimes) do work with GPS data and 3D orientation on an engineering side (i.e. relatively far from the mathematical side of it). Isn't the great circle effect inherent if you are not using a 2D representation of a map? My layman understanding is if you calculate your destination / origin based on sun angles etc, the solution would be inherently represented by great circle route (This coming from a point of poor knowledge on my behalf, not making any statements).

Click to expand...

A key distinction here is between a great-circle route and a rhumb line, and it's very easy to conflate them. That's in part because they look pretty similar on non-oceanic scales.

A rhumb line is the route you take if you follow a constant bearing with respect to (most typically) due north. On a globe, it ultimately tracks out a spiral around the pole.

In contrast, a great-circle route will constantly change the bearing (other than in special cases like travelling directly north or south, or travelling along the equator): for example the route LHR -> LAX starts off going north-west and ends up south-south-west. This route is directly ‘pointing at’ LA from London, which is obviously true on a globe, but looks wrong on most 2D map projections. That is to say, a great circle is a ‘straight line’ on the surface of a sphere: if you start off in the right direction and ‘keep on walking straight ahead’, then what you trace out is a great circle (it's initially surprising that this process is not what you're doing when you follow a rhumb line). This path is also the shortest distance between the start and end points.

The technicality that's special about the Mercator projection is that if you lift a bearing directly from that map, and sail constantly along it, then what you will trace out is the rhumb-line route from A to B, which will get you where you're going, though not quite by the geographically shortest route. If you were to take that bearing from a different map projection, you'd end up in the wrong place.

Going from start- and end-positions as latitudes+longitudes, and calculating the great-circle connecting them, is one of the standard exercises when astronomers or navigators learn their spherical trigonometry.

Bonusround said:

In the same vein there is also Steamac – an ARM-native build of SteamOS, from the Steam Frame, running virtualized on macOS.

Let the benchmarks begin!

Click to expand...

Ok, I was wondering when something like this would show up.

doraemon said:

Definitely felt like a heavyweight match, with neither side able to pull away. Very low-scoring, which is not too surprising when both teams have excellent defense.

Click to expand...

Defense in a pro-basketball game? Unpossible!

Taunted Happy Fun Ball said:

Using that approximation, if your typical American car weighs two American tons, this thing does 26 times as much road damage.

(Or over 446,200 times as much as a fat man on a freakishly heavy bicycle.)

Click to expand...

The Netherlands tax cars on the basis of their weight, but not applying a fourth power multiplier, which would have been appropriate (should it be weight of footprint pressure?). But they never made any mass market cars (other than the belt drive DAFs) so they have little control over what was available on the market. Here's the link to calculate your tax liability: https://car.thetax.nl/en?provinceKey=NH&fuelType=Benzine&volume=1551

He was the best. We need more like him.

Madestjohn said:

“Elon Musk is a genius, a true genius,”

Putin
Oct. 28th 2027

Click to expand...

Both of them share responsibility for making it so that Ukrainians need killer drone tracking apps the way people in the US/EU need thunderstorm tracking apps.

Just out of curiosity, as I haven't participated in the WoW thread in a dog's age...

Forever talk okay being mixed in here with the Retail (and presumably the multiple varieties of Classic)? Or should it migrate to a new thread after it launches in a few days?

Cthel said:

Ukraine release an app that shows near-real time tracking info on Shaheds

Spoiler: Telegram Embed (UKR)

Login to view embedded media View: https://t.me/kpszsu/92022

I wonder how much discussion there was over the need to balance accuracy & timeliness against not giving Russia an easy way to identify holes in Ukraine's Shahed-tracking network...

Click to expand...

It's fucking horrible that in 2026, Ukrainians need an app to track killer drones from Russia the way people around the world need an app to track thunderstorms.

JohnDeL said:

Given that the ground based telescopes have areas 17 - 150 times larger than the Hubble, they don't generally need the long dwell times.

Interesting how you shift the goalposts from sharpness (which ground-based telescopes now do better on than Hubble) to dwell times (which ground-based telescopes can approximate by summing exposures, much the way that Hubble did for its famous "million second" exposure).

Click to expand...

The goal posts are, and always have been, can ground based measurements replace any measurement that Hubble can. Showing examples where ground measurements can replace individual Hubble measurements is not, and never will be, sufficient to show that. (I'll add that talking about Starlink and other upcoming constellations is valid as well. That is the environment a boosted Hubble and the ground based telescopes will be operating in.)

Talking about a scene requiring long dwell time is a stand-in for imaging a very dim scene. Pretty much the opposite of the first image you offered as "proof" that ground based telescopes can do anything that Hubble can do. That was obvious from my first post.

A particularly strenuous example would be something like Hubble Ultra-Deep Field. That will still need a long dwell time for ground-based observations. Not a million seconds, but it will need a similarly large factor more observation time, as compared to observing Neptune, that Hubble did. Talking about dwell time is absolutely fair. Maybe ground based observations can handle a similar case. But you haven't offered up anything like that.

boost it to orbit around the moon so it can keep an eye on the chinese base there

solarbonite said:

So there's a market for "GPS for Birds?"

Click to expand...

Or wing-powered self winding watches so they can check the actual time.

redleader said:

This feels weird but for the first time since like 2016 I'm actually really optimistic about intel. 18A looks great and as far as we can tell it's actually ramping close to on time, which intel hasn't managed to do with a leading edge node in 15 years. This RAM disaster is obscuring everything in CPU space, but at some point the bubble will pop and we will be back to having two competitive CPUs at the same time.

Click to expand...

We’ll see what happens but most of the current rumors are Intel’s Coyote Core P-cores don’t clock very well, so AMD should have a clock speed advantage this gen, and 50% more P-cores (12 vs 8) on a single chiplet. If true, Zen 6 X3D should be the easy pick for gamers, games that use a lot of threads can all sit on a single core type, and should run substantially faster on AMD. Also Intel’s past crutch of relying on fast RAM (8K+ MT/s) to be somewhat competitive is probably going to be an additional downside given DRAM costs. It may be Intel’s main niche will be the 52core variant for those who want/need a lot of multi-core performance this gen.

IceStorm said:

Bartlet Lake has a 12 P-core, no E-core variant. That's the impetus. You can have 12 P-cores at 4.9Ghz in Socket 1700, with hyperthreading. Arrow Lake and Nova Lake don't have an option like this. I think it's... Titan Lake? It's the thing after Nova Lake where we'll see hyperthreading return to P-cores.

Click to expand...

Isn’t that Hammer Lake? That should be the last CPU gen on LGA1954 which brings unified core back to Intel so it’ll be “all P-cores” again.

TheBaconson said:

Point 2.
Where is this premium market? They weren’t around near the end days of Concorde. The crash was just the final nail for them, they were loosing money for years before, because the market was not there anymore.

Click to expand...

The premium market that wants to save ~2-3 hrs US East Coast to Europe (or vice-versa) can more cheaply do so flying private or charter. Flight time is not travel time. The combination of business jets with intercontinental range AND the ground security and customs time mean that a subsonic business jet leaving from the private terminal on your schedule will be a net shorter trip. It only gets worse when the business jet lets you fly directly to your destination and avoids a layover. Worst case, there's a re-fueling stop, which is still shorter than a traditional layover.

Additionally, an perhaps more important for the airlines, a 1.5x faster flight does not close the use-case for US-Europe flight schedules. The current flight times and schedules between the continents have been hyper-optimized to ensure the aircraft flying the route arrive:

• In time to make connections (at least hub-to-hub)
• Avoid nighttime takeoff and landing restrictions - applicable at a lot of European hubs
• Can complete a turn and get back in the air before the restrictions

Modern widebodies (777, 787, A350, A330NEO, etc) flying these routes get turned ASAP and can complete a back-and-forth from US to Europe over and over with ground time dominated by the turn time, not airport restrictions or alignment with connecting flights (that are driven by 'domestic' schedule).

While a smaller aircraft can be turned faster (potentially a lot faster), at 1.5x faster speed the flight time it still won't be short enough to get 2 round-trips a day. That means the airframe is sitting on the ground earning nothing... and that's a huge problem for a massive capital asset. This wasn't a problem for Concorde because Concorde's economics really didn't matter (and the biz-jet competition wasn't a thing)... but that will not be true for a modern supersonic airliner.

nxg said:

...

The other demonstrations of statistical power included that the observers were able to rule out the birds having available corrections for the Sun's altitude, or having a magnetic compass; and they were able to determine that the birds are following great-circle routes, as opposed to a simpler rhumb-line route. That's pretty spectacular – birdbrains doing spherical trig!

...

Click to expand...

Help me out here, I (sometimes) do work with GPS data and 3D orientation on an engineering side (i.e. relatively far from the mathematical side of it). Isn't the great circle effect inherent if you are not using a 2D representation of a map? My layman understanding is if you calculate your destination / origin based on sun angles etc, the solution would be inherently represented by great circle route (This coming from a point of poor knowledge on my behalf, not making any statements).

flyxnite said:

Extrapolating the benefits of a turboshaft application to a turbojet application is quite a stretch. Completely different way to harness & control energy; pounds of thrust versus pounds of torque. Add the utterly different operating environment of these powerplants & accessory drive requirements, they become completely different animals. Ask any pilot that's flown both turboprops & jets.

The Pratt & Whitney PT-6 turboshaft engine is an example of both ground (remote pumping stations) & aviation (ubiquitous turboprop engines) applications but jet power it is not by any stretch. That's some bullshit marketing/investor speak.

Scott Kirby isn't a stupid guy...

Click to expand...

Jet engine core are also used as ground power. The old CF-6 of 747, DC-10 and A300s was a common enough conversion. Whip the fan off and use the core as a gas generator or somehting like that. I know they were used in ships. You get the opposite happening too. The engines in the BAe 146 are the ones out of the CH-47 Chinook with fans (and a small gearbox) added.

I would also add that the PT-6 also saw use as an F1 and Indycar engine in the back of the Lotus 56 though for ground use it's normally referred to as a GT-6.