I think if you get over 3 years from when money was conveyed it is more difficult for bankruptcy court to claw back money. If the patent trolling isn't sufficiently lucrative, Blackberry gets to keep the $450 million and leave the lender holding the bag. Also, being a creditor in the bankruptcy might give them some leverage also.

Scare quotes on fraud because it might not technically be legal.

June was a local minimum for crypto. And, all the coins seem highly correlated, so hard to say what "trading" is supposed to do. So, iof course the hamster did well buying in a trough. If the hamster has started May 5th, I bet it would have done terrible.

From a cost stand point, Uber and Lyft can never compete. The costs of hire car are car purchase, maintenance, insurance, and drivers. Uber and Lyft put all those on individual drivers. The thing is fleet management and maintenance has economies of scale, that Uber and Lyft's individual drivers can never touch. So, the only option for them is to squeeze that additional cost out of the drivers and somehow also squeeze out a profit for themselves. Of course, when they squeeze drivers too much then people won't drive for them, and then riders can't find rides and then they don't make money. The only way they could become profitable is to get a monopoly and jack up prices. Unfortunately, their business plans involved eliminating the barrier to entry that would allow them to have a monopoly. Basically, Uber and Lyft can never make money with their current business model. Driverless was the only possibility where they could eliminate the driver and get the fleet management economies of scale. Although, that was most likely not serious and more about hype to get to the IPO and keep the stock price up.

But, they got their IPOs, they should be doing everything they can to keep the stock price from tanking, so all of the pre-IPO investors that financed the billions in losses can get their money back from the rubes who think these companies can ever be profitable. They will be the ones left holding the bag when Uber and Lyft run out of cash.

At the current rate my math says that there will be enough doses by the end of 2022 based on the article. Unless there is a word missing from the article. The article says 11 billion doses are needed. There will be 6 billion doses this year which includes the ramp up phase. The average production rate this year is quite a bit less than the current production rate. So, producing another 6 billion doses next year is pretty much guaranteed which makes 12 billion doses by the end of 2022 which is >11 billion. Distribution and economic ideology are probably bigger impediments. That is unless the missing word is "more" as in "11 billion more doses are needed".

The protein sub-unit vaccines (Novavax, Sanofi and GSK) that are in phase 3 right now are probably a more interesting factor in changing the production and distribution issues. Protein sub-unit is not a new vaccine technology like mRNA so should be manufacturable at more existing vaccine manufacturing facilities. They also don't require the super cold refrigeration of the mRNA vaccines, which should help with distribution to low income countries.

111F ~44C on Sunday on my outdoor thermometer. The most uncomfortable thing is that the low last night was not supposed to be much under 80F ~26.5C. So, you don't start today as low as you would like inside to get through one more day. Luckily at my house, I am at the bottom of a valley and it seems to have cooled off more last night (67F ~19.5C) than the nearest towns that are not in the valley.

Ursine mammals defecate in sylvan settings.

Yep. Bad science reporting. They author clearly decided to override the source article and replace the formal term hypothesis with colloquial theory.

I also checked the NEJM article. 5 uses of hypothesis correctly and 2 uses of theory. One of where it is used to refer to a "long standing theory of viral pathogenensis", which seems like correct usage. The second use is ambiguous due to pronoun trouble. "this theory" might refer to the preceding theory of pathogenesis, but could also refer to the hypothesis.

One reason their is a big loop is to accelerate the particles to huge speeds/energy as they zip around the loop. With satellites there would be no acceleration between satellites.

Part of the reason bigger loops are needed for bigger energies is that we are talking about charged particles. A charged particle at a constant velocity in a circular path is accelerating and an accelerating charged particle emits energy. The tighter the loop at a given speed the greater the acceleration and the greater the energy lost. That is in addition to being able to accelerate particles as they circle the loop over and over. As compared to a linear accelerator where you wouldn't have to deal with curving the particles, but get one shot at acceleration.

I was thinking they screwed up their accounting badly. Now they are going to have to pay the actors they were trying to screw over with $0 up front and net points on the back end.

Which makes Wakanda not a terrible choice being fictional, since it is fairly easy to notice, and no one is likely to use it to make real life decisions. Using a real country with fake test data would be harder to detect. I might have used "Afaketestcountry", "Bfaketestcountry" etc... but that wouldn't be as fun. I wonder if there is test data for Latveria.

I agree with you about design patterns. If you are not using design patterns (regardless of language) you are either reinventing the wheel and probably doing it worse than you could have because you have ignored the accumulated knowledge of those that came before you. I heard somewhere something along the lines of; Any developer should know their design patterns. A skilled developer knows which one to use when.

In addition, they help with communication, which is critical for large projects. If everyone knows Strategy Pattern, then saying something uses strategy pattern explains a lot if very few words.

I think the problem the author is describing as nothing to do with languages or OOP, but the idea that experience and skill can be replaced with any rando with a degree or certificate.

Uber and Lyft also have the problem that fleet management has economies of scale that Uber and Lyft cannot realize since individual drivers own and maintain their vehicles which do not have those economies of scale. So, basically, Uber and Lyft drivers have higher cost than other hire care services. So, Uber and Lyft have to pay enough for the individual drivers whose costs are higher than taxis to at least seem like they are making money otherwise drivers will stop driving. In order to try and create a monopoly they have to charge less than the taxis that have lower costs. So, they will lose tons of money. On top of that Uber and Lyft are destroying the barriers to entry that the taxi services have. So, if they get the monopoly and raise prices to where they are actually profitable with their higher costs, new taxi services will spring up and undercut them since they have destroyed barriers to entry in the hire care business.

So, Uber and Lyft need to basically destroy the barriers to entry, so they can put the old hire car services out of business and get a monopoly, then get local governments to put those barriers to entry back in their favor. And, we will have a higher cost business charging everyone more. Enjoy your Silicon Valley Venture Capital subsidized hire car service while it lasts because it will either start charging a lot more or go out of business.

Self driving cars is a Hail Mary in the hopes that it might somehow justifies the billions of venture capital money that Uber and Lyft have burned.

Coming from a Spring Java application, that did migrate off AWS. The problem is not the application migration. You have to provision and configure the cloud services (VMs, databases, networks, load balancers, etc...) which uses the cloud provider's APIs to create. And, if you are being cloudy, you do that using some kind of automation which is written using the cloud providers APIs. So, the majority of the work is in rebuilding the infrastructure on which your application runs in the new cloud provider.

The thing is I cannot think of anyway a platform migration doesn't impose costs that will feel like lock in to whatever current platform you are on. Cloud Foundry or Kubernetes/Docker might reduce the cloud provider lock and being open source with multiple vendors at least the platform lock in is not a vendor lock in.

Google's work on HTML frameworks like Angular where a lot of HTML in the DOM is generated by the framework from some other abstraction, makes me think that the change had nothing to do with Microsoft. An upgrade of whatever framework they use for YouTube could have inserted the empty div. Which might appear useless in the particular context, but if a different part of the framework were used it would not be empty.

Basically, the framework trades off what should be a harmless empty div in some cases for ease of implementation of other cases where it wouldn't be empty.

Now taking advantage of that to release benchmarks is suspect.

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FDiet_drink%23Amount_of_artificial_sweeteners_in_diet_soft_drinks

For what it is worth.12 oz can is about 350ml. It looks like 150mg is the top end for artificial sweeteners in a can of a diet drink. So, about 0.5mg/ml or about half what was used in the study.