It's not clear that will ever be needed. EV batteries don't just stop working (barring some unusual fault); they just gradually decline in capacity, and the decline is very slow after the first 1-2 years. So expect to have 95% of capacity after two years, 80% after a decade, 60% after two decades, 50% after three, etc.

So it's just a question of when the capacity drops so low that the vehicle no longer has enough range -- but over time charging infrastructure is going to get better and better, so long range will become less and less important. Also, batteries are going to get cheaper.

So, yeah, it seems entirely reasonable to me to replace the battery in 20 years (if you haven't replaced the vehicle by then). Especially since the fuel savings over that time will far more than cover the replacement cost, even if the replacement cost hasn't come down, which it will!

That was 12 years ago. A 12 year out of date critique of a web technology that has had ongoing language updates and two entire rewrites in that interval should be viewed with some suspicion. Also, are you really just citing the title of the article and none of the content?

I'm not even defending PHP here, just questioning lazy kneejerk, "but it sucked once, so now I hate it forever" thinking.

Turbotax offers free service to low-to-moderate income people as part of an agreement it has made with the IRS. In return for this, the IRS doesn't provide free electronic tax preparation services like most other advanced countries do. For most consumers, the IRS could in fact automatically fill out their returns and the consumer could simply check it by answering a few simple questions rather than puzzling over instructions written for professional accountants.

If you've always wondered why filing your taxes couldn't be simpler, a bit part of this is marketing from companies like Intuit that make a lot of money out of simplifying the process for taxpayers.

The free tier service is something Intuit is contractually obligated to provide. Upselling low-income people to a paid service that wouldn't benefit them in any way is morally dubious at best.

What I wonder is why flywheel storage isn't a popular alternative to batteries for fixed installations. Creating heavy flywheels isn't hard, nor does it require the kinds of toxic materials used in batteries. Tungsten carbide with a steel casing anyone? Mount them on good bearings and you should get >90% efficiency from them, and I'd guess the energy density should be higher as well. I'd have to do some calculations to see if units small enough for home installation would have enough capacity to be useful for a reasonable length of time, but large-scale industrial/commercial installations should be.

...this was the third repetition of the same basic scenario: issue is known and known to be a risk to crew survival, issue occurs several times without anything happening, NASA managers decide the fact that it hasn't caused a problem yet means it'll never cause a problem, issue occurs again and causes a problem resulting in loss of vehicle and crew. One of the first things Scarne says in "Scarne on Cards" is advice for gamblers: the odds don't tell you it won't happen, they tell you how often it will happen.

Note: Boeing management is dealing with the results of that same scenario playing out yet again.

The exponential growth achievable by Von Neumann probes that don't mind traveling for millions of years to get to the next galaxy make space small.

What are these scientifically-sound indicators?

I picked up my Klein Stein at Cliff's house a number of years ago. He's a delightful character and we veered off into conversations about slide rules and oscilloscopes (he has a large collection of both) as well as computers in education and a number of other topics. A number of his large bottles were on display at San Francisco Airport.

There's usually more than one way to do it, yes, but more often than not there's only one or a very small number of "best" ways to do it and most competent programmers will immediately go for those. For instance, you can iterate through a list/collection type using indexes and a for loop but nobody's going to do that, they'll use an iteration construct (foreach loop or Each() method or an equivalent). Idiomatic code is a thing, and usually the concept makes sense to non-computer people.

Training AI to catch someone whipping it out is easy. There are entire data centers devoted to well curated/classifieds and properly tagged porn.

I'd bet you'd find it is some of the highest quality training data out there if you want classifications of nudes or sexual images.

No one want to sit can tag various forms of busses or bicycles in their captcha ... but rest assured am army of basement dwellers and republican senators have ensured that your dick picks and labia are well organized and tagged for the good of their fellow man

That's the 'girlfriend experience' right?

Really what's happening here is Adobe is building AI to replace OnlyFans models.

OMG you should definitely patent/trademark/copyleft that or whatever, the upfront investment will yield billions in return if they can have the AI chat and show video to the customer based on AI generated content. The weird extra/missing finger thing will even be a surcharge feature for the fetish crowd.

You sir are ahead of your time.

The code in question wasn't ever under MPL, it was added by Hashicorp after the license change. Which doesn't matter much, if you compare that code to OpenTofu's code the structural and cosmetic differences are obvious. Note that the articles only link to the code and don't actually show it side-by-side, probably in the hope that people will take the claims at face value and not investigate whether they're correct or not.

I compared the two sets of code mentioned. IMO the OpenTofu code is not copied from the Hashicorp code. There are structural differences that point to the OpenTofu code being independently written, along with the obvious differences in variable names and such. The claims seem to me to be a case of someone trying to smear a successful fork to keep it from gaining traction.

I suppose some of that may be down to the difference in the value of the change. It was worth about 2.5X what it is today back when I was working convenience store night shifts, so people might have cared more about getting it correct.

Even more, people at Starbucks are paying $7 for a cup of coffee, so they're clearly not very price-sensitive. If the customer doesn't bother to look to see whether they got the correct change, should the cashier waste everyone's time getting it right? I think yes, but I could see where people might disagree.

I know people at the convenience store got pissed when they got shortchanged, which is why cashiers who couldn't count change out got fired pretty quickly. They might last longer at Starbucks today. Especially since most customers don't pay with cash.

That's not new. I saw the same problem when I was working at a convenience store in the late 80s.