Your right about everything you said, except for the bits about nukes. You must live in TX, so you know about STP. We could just clone that design and built a couple dozen more its only 30 years old (although the design is still from the 70's). It took 13 years to construct the first time around, surely we can do better the second time around if we don't change the contractor a half dozen times and go through a bunch of political BS to slow it down.

Just a two copies of that plant would be roughly the same power output as the entire installed wind capacity in TX. 10 copies and we could replace all the NG in use. According to Wikipedia it cost $15B in 2019 dollars to build that plant including all the contractors being replaced and all that bullcrap. So for roughly one year of the texas budget we could wipe out all the remaining coal plants and over half of the NG plants in TX, which would result in effectively 100% carbon free energy production and we probably could do it in under a decade like the french if we wanted to.

Your are looking at MW/H costs for nuke vs wind.

Right, just like everyone else.

You can't compare the $/MW of wind vs nuke because they aren't the same the wind power is missing a huge external cost. They aren't the same because the wind is intermittent and you have to add in the price of the NG plants that are running for some percentage of the time. So the minimum cost is the NG plants you need to build to back them up. AKA the minimum price is the average of the NG plants and the wind plus whatever you cost you want to assign to the remaining externality of the CO2 the NG is pumping into the atmosphere.

AKA your computing your green energy prices wrong to lie about their costs. And the CO2 producers love it because they can green wash their projects with the couple wind turbines they install over 20 years (like in TX) while the energy mix goes from 10% reweable to 20%. In another 20 years TX might be 50% renewable, but the total carbon output will have gone up because the total energy production quadrupals again.

Yes, i'm rounding in TX, but the point is that while the wind rollout is huge, so is the energy demand growth. Last I looked the green energy rollout wasn't even keeping up with the demand increase. Which isn't going to decline as long as people keep moving here and building crappy houses.

Nukes take a decade to build because id10t green energy/etc people NIMBY them.

Battery+Wind+Solar rollouts _WILL_ take longer than a decade too.

So if one stands here today, and says how do I solve this the fastest, the answer remains the same as it has been for the past 40 years. Build 100% nukes, once that is done we can talk about whatever magic green energy solution is going to solve the problem, be that some new battery tech or some breakthrough in energy transmission, whatever.

Please, this comment could have been from 1970s, and its 100% of why we this problem hasn't been solved.

Texas has one of the largest green energy rollouts in the world, and its a failure when it comes to reducing C02 emissions. For every MW of green energy installed we have 3X in NG load following. Sure for a couple days a year, we are 100% renewable but for the remaining ~300 days a year or so it can get so bad that we are at 100% CO2 burning.

To solve this problem, the green energy people need to shut the * up if this is all they have to say. To solve this problem what we need are a 1970's level of nuke rollout like France. 100% nuke, and we encourage people to buy electric cars which are themselves load following and only charge during non peak times. Then instead of decreasing production when there is excess supply it needs to be used for one of the many C02->liquid fuel recapture cycles for use in air travel/etc.

Yah, its beginning to irritate me two. I live within the city limits in one of the largest couple dozen cities in the USA. I have a single provider, which charges > $120 a month for 400/35 service and only upgraded from 20/1 a few years back when it looked like google fiber was going to steamroll them. If I lived in the part of town with google fibre, I would have the choice of three different providers, two of which offer symetric gig, and one which offers "gig" service with cable modems for less than $100.

The rural townships in a couple hundred mile radius frequently have fiber drops as well, usually for far less than $100 because they used the universal service fund money to string fiber in their towns.

Maybe instead of the rural/urban divide we start ignoring the stupid FCC maps which list the best internet possible in a given census map that may only be available to a single person and start looking at what the _worst_ available in a census map is, and try to bring that up.

Also, it needs to consider upload bandwidth. If someone can't use crashplan, or zoom because their stupid provider allocates all the spectrum to download bandwidth (or to keep people from providing their own content, or whatever) then it doesn't count either. For purposes of determining the "slowest" internet it needs to be the slowest of upload or download.

Your right, but 32-bit support in windows was released in 1993 after a long beta for both NT and win32s, which at the time of this posting is nearly 28 years ago. I'm guessing your counting from win95 which wasn't the first 32-bit windows.

https://en.wikipedia.org/wiki/...

Yah, how much do rental trucks that can tow a camper cost again this week?

No, I have a single vehicle I use for multiple things.

Suse studio was probably one of the nicest things in the Linux market, fully automated image/installer creations for custom appliances/applications. The problem of course isn't that they weren't great, it is that they likely had a hard time monetizing it so it stagnated due to lack of investment. That describes a large part of the Linux ecosystem. 80% projects where someone spend a few months hammering away at something, didn't make any money but it fills a niche so it zombies around in a half-ass shape full of bugs/etc for years because the original author moved on to something else, and no one wants to put in the hard work of maintenance. The successful opensource projects overwhelmingly have large corps behind them funding the development for reasons which are fundamentally selfish.

Well the CADT model is alive and well in Linux. People who write and use civil engineering software could give two shits about the latest gnome rework to use web technologies, or the latest rework of the sound or init system, much less the wonderful perf improvements of btrfs that come with complex and arbitrary rules about actually persisting the data to disk in case of power failure. Those people want to wake up in the morning hit space and have their computer resume to the exact same engineering diagrams they have been waking up to for the past decade.

MS built their desktop/workstation market on three decades of absolute backward compatibility and slow and slight changes to the OS and UI (that were almost always revertable with simple registry flags). Now that they have decided to remove one's ability to control the button bar, or whatever random UI bullshit every two weeks, maybe there isn't a difference but as long as they continue to support applications written for win32 25 years ago, those companies have little interest in porting to linux with their engineering time, rather than bolting on application specific features.

RH and to a lesser extent Suse exist to create a bridge between the CADT model and one that is boring where your server just keeps chugging along running SAP or whatever day in an day out. But the OS surface for server applications is like 1/100th that of a desktop application.

That tends to be true of most of these EV's the range is beyond what most peoples daily driver/etc is.

The problem is road trips, which I take I'm a bit of an oddball driving across texas a few times a year to escape the shithole area I live in 700+ miles of driving at 80 miles/hour per day for a day or two.

So, it remains to be seen how these trucks behave if one puts a couple thousand pound trailer behind it. Be that a flatbed hauling construction supplies, bobcats, etc, or just boats and travel trailers. A normal ICE truck that can pull off ~23-27 mpg without a trailer frequently halves or worse that pulling a trailer.

So I would expect that if the truck says it has a 350 mile range to be half that when pulling something.

Besides the shitty UI, those touchscreens are likely just another form of planned obsolescence. In 15-25 years, how much is the car going to be worth vs replacing that touchscreen? Given that ford regularly charges > $200 for simple modules with little more than a a couple FETs/relays with a bit of 74xxx series logic, and the ECUs are a few thousand, I'm betting that screen is a thousand or three to replace. Which is a hard sell on a 15 year old car that's worth 5-7k.

Really inefficient native code because the java bytecode model isn't really designed as an IL for modern register oriented machines. Further all the dynamic functionality has to either be trapped and interpreted or turned into pretty inefficient native code to deal with the dynamic nature. Plus garbage collection/etc tends to fragment up the heap/etc so one gets poor memory placement, which destroys cache locality, etc, etc, etc.

This isn't really just a java thing, python has similar problems without all the decades of heavy lift optimization work that java benefits from. So, yes java/php/javascript easily trounce python and most of the other dynamic languages. OTOH, when compared with the common static ones c/c++/fortran/rust there seems to be a rough upper bound on how fast they can get, with the static ones tending to perform better with far less optimization effort. Most of the language benchmarks will use -O2/3 or similar for gcc, but forget to -march=native or -mavx2 which for a lot of problems can easily double the gcc/llvm perf.

INAL, but I neither admitted to anything, nor do I live in a state where the buyer is responsible in a situation where the object was purchased outside of the state I reside in, much less outside of the country (as is the case here afaik). Which is generally regulated by the feds (yes there are loopholes around liquor/etc). The newegg decision is when all the amazons/etc started charging sales taxes cross borders, and its still unclear for many transactions.

So, to claim such things are silly.

Looking at my shopping history seems I've purchased a number of aukey chargers and cables. Frankly, thinking about it, they have been some of the better chargers/cables i've purchased. I've only had one cable fail (it started smoking) while hoards of garbage from local retailers (I tend to buy the cheapest cable on offer) I've purchased has failed.

I wish they were more explicit and had tools other than ban hammers if the problem is fake reviews.

Although on the plus side, it looks like the aukey site itself is a few bucks cheaper than their amazon store was (there are promo codes to reduce it even more), and has the advantage its not charging sales taxes.

I guess it doesn't have reviews either, but everyone knows that amazon is full of fake reviews because that is part of the gameification of getting ahead in these marketplaces.

Unless it was something truly nefarious amazon is probably just shooting themselves in the foot because the brand is probably large enough to survive on its own now and people will just buy direct.

Have tsmc/samsung/intel been doing press releases for every research/process tweak that they have been doing? I don't get that impression, tsmc is talking about volume production of 3 & 4 nm in the next year or so. That means they probably produced the first chips on the process a couple years ago and they have spent the past couple years perfecting it so it can be produced in volume.

So really this sounds more like "we are way behind, our competitors are about to do volume production on 3nm and we just got out first 2nm working, so we might have it in volume production 2-3 years after they do"

For ref:
https://www.anandtech.com/show...