No argument there. Bitcoin is sometimes touted as a micro-transaction system, but in reality it isn't really scalable even as a macro-transaction system.

I could see how it could evolve into some kind of large-scale currency exchange/basis/reserve system that other more scalable micro-transaction systems become based on. So, bitcoin controls the overall supply of currency but all the day-to-day happens elsewhere. I'm not sure how that would even work.

However, the idea of the average person doing average transactions in bitcoins just can't work with the current design.

The distributed nature of the system also can't handle simply increasing the supported transaction rate. You can do it, but eventually you need to be the size of Visa to handle all the transactions.

The distribution of bitcoin isn't about splitting up storage of transactions across many parties. The distribution in bitcoin requires every party to have a copy of every transaction. That obviously created fundamental scaling issues.

Part of me wonders if the right solution is to increase block size, or to increase block rate. If the effort and reward for blocks were reduced, then you'd have less time between blocks, which still gets you a higher transaction rate, but it also gets you less latency.

What I don't understand is the relative impact on communications, processing, and storage of all those blocks.

There is a bit of a fundamental limitation on the nature of bitcoin. Right now the whole model of bitcoin relies on clients maintaining a full history of the entire blockchain. That is a bit like saying that to process Visa transactions you need to maintain a record of every Visa transaction that was ever performed since the beginning of Visa, and maintain it in realtime. Ie, to use Visa you have to be Visa.

That is going to be a big limitation on mainstream adoption of bitcoin.

Well, hopefully the FCC's case puts an end to these practices.

Otherwise I could see somebody writing an app for any device whose wifi can be put in monitoring mode that just sends a de-auth frame for any Smart City connection it spots anywhere. Then nobody has wifi at their conferences and they go out of business.

That is certainly possible, but I suspect it isn't the case.

In order to get something to you there are a few steps:
1. They have to have it in inventory.
2. They have to package and ship the item out.
3. The courier has to deliver it.

#3 should be the same for everybody, though I'm sure Amazon can negotiate lower rates on the basis of volume and integration (it isn't like the guy at Amazon picks up the phone and calls UPS to ask for a pickup, etc). Also, Amazon can save money by keeping lots of metrics on delivery times. If the USPS gets stuff from zip code 1 to zip code 2 in 2 days 99% of the time, then they can sell that as 2-day shipping and bribe the customers with incentives anytime there is a miss, rather than paying 2x as much for guaranteed 2-day shipping from somebody else (which also has some failure rate anyway).

A large hardware chain could probably achieve #1, but whether they do is debatable. You'd be amazed at how poor big companies can be at inventory management, and historically this has caused many a company's ruin. Walmart killed KMart and everybody else largely on the basis of really well-managed inventory. Dell nearly killed just about everybody else back in the day because of just-in-time inventory management for rapidly-depreciating assets like CPUs (you can't buy a $500 CPU and then take a month to sell it when your competitor just buys it a month later for $400).

However, I think #2 is the real killer. Sears and Amazon may very well have the same product available in their warehouses. However, with Amazon after you click the button they may very well have a robot with the item heading to the packing line in 15 minutes, with the shipment data already transmitted electronically to their carrier. They are extremely efficient at getting stuff out the door. An item that somebody wants to buy and which is sitting on your shelf is just a waste of space and money, and customer satisfaction as well.

In my experience that is where most companies fail. With Amazon if I buy with 2-day shipping the thing is out the door same day before 5PM or whatever (and that cutoff is very late in the day compared to many competitors). With Amazon 2-day means that I have the item in 2-days. With most other big vendors if I buy with 2-day shipping it often means that they take 3-4 days to ship the thing out, and then it arrives 2 days later. So, I'm paying that premium on shipping just to watch the vendor fumble around with my order, and I never really am sure about when it will arrive.

Amazon is pretty ruthless on this stuff, and IMHO their practices are so ruthless they border on human-rights violations in their actual warehouses. However, even if they cleaned that stuff up they'd still be way cheaper and faster than just about anybody else. Companies that want to compete have to invest a lot more in streamlined processes, because customers aren't going to pay for mail-order that takes a week to arrive.

Isn't that a bit like Walmart displaying yesterday's prices on the store shelves and on their advertising, and you find out what your real bill is after you hand them your credit card? But, for a modest fee they'll show you today's prices instead?

I get that commodity prices change in realtime. I don't propose that when you are quoted a price that it be valid for a day or anything like that. However, I see no reason that anybody shouldn't just be able to see what the price of a commodity is at any time, accurate to the time at which the price leaves the webserver or whatever.

The problem is that none of these are useful services on the scale that HFT does them.

If I got to sell a stock and have it sold in 10 seconds, I'm pretty happy. Market makers and such which make that possible do provide a useful service to the economy, versus me having to sit on my order for three days until somebody else notices it.

On the other hand, being able to sell it in femptoseconds instead of microseconds doesn't add real value. It just becomes a pay-to-win scenario.

Likewise, keeping international markets in sync so that the price isn't $5 in NYC and $10 in Tokyo is a useful service. On the other hand, ensuring that when the price goes up by 0.01 in NYC that the price goes up in Tokyo by 0.01 within microseconds of the speed-of-light travel time isn't a useful service. It just doesn't have to be that close.

That doesn't make any sense. Why would you regulate a corporate-run service that vets its drivers and passengers with centralized realtime tracking in the same way as a bazillion completely independent taxi companies that keep all their records in the car and rely on the drivers for accurate recordkeeping?

I think the ultimate end-game is a set of different regulations that do apply to everybody, so that taxis and Uber run under the same rules. I don't have a problem with independent taxi companies doing things the uber way as long as they're large enough to have skin in the game (ie every car isn't owned by a separate holding company or whatever), and continuously track their driver performance and have realtime monitoring of cars and who is in them (both the drivers and the passengers). That is what Uber is doing, and that is why people want to use Uber.

People think this is about saving money/etc. Sure, on some runs with flat fees or whatever or when the driver is gaming the route that might be the case. However, I think most people would rather deal with a big company that has one big reputation to protect than a random cab company that really doesn't care what you think of them since you'll probably never see them again. When I travel I use Uber in multiple cities, and it is the same brand. If one of their cab drivers kills somebody, their entire international business will suffer. They have incentive to police themselves. The same is not true of "Lucky Cab Holding Co" in NYC that owns a single medallion.

It is no different from hiring a plumber. I can hire a local guy and save $50, or I can call a national chain and know that if I complain I'll probably get taken care of. There isn't anything wrong with having both models, but with taxis to date operating 100% under the "local guy" model it shouldn't be a surprise that there is demand for something different.

And I'm not convinced that Uber will ever take over 100%. I'm sure they'll reach some kind of market share balance with traditional cabs. After all, if I am at the train station there will be a line of cabs waiting, and Uber might be 10min away. Both models have some advantages.

What's the difference? The way you challenge a law is by acting as if the law doesn't apply.

It actually is a lot harder than you think.

EMP has to be REALLY powerful to do something like fry circuits, and you can imagine the havoc that could cause in general (that drone is far away, lots of other computers are a lot closer even if not being aimed at). I don't know how long-range you can even direct EMP generated using conventional means - the stuff that wipes out cities is basically a byproduct of a nuclear explosion (as far as I understand it this is basically just synchrotron radiation from ionized air).

A GPS jammer is also hard to do since there is a simple countermeasure - put foil around the underside of the drone. The GPS satellites are above the drone, your jammer is below the drone. Your jammer is also going to cause a lot of problems with GPS systems all over the place.

Next step - drone drops cutters to cut through the nets, or something to burn them. How long before people start sticking guns on drones to take pot shots at the guards? If done from enough altitude so that the drone or muzzle flash isn't easily spotted the thing is going to be invincible to anything short of radar-guided AAA (which might just be radar-driven shotguns, but still).

With decent airflow (no windows left, building is hundreds of feet tall with no horizontal obstructions at that level), and enough temperature, virtually anything will burn. The airflow and temperature are competing factors to some extent, but I'm sure that fire got quite hot.

It isn't like they mail the prefix to you and have you hand-type it. They're handed out by NDP, which effectively works the same way as DHCP - the ISP will give you a prefix.

Even if they don't need to change prefixes, they still have incentive to do so. If you want a static prefix you will have to pay them more. If you don't care, they can give you a new one frequently.

That is what most ISPs do today with IPv4. Why wouldn't they do the same with IPv6? There is no technical reason why ISPs couldn't give all their customers static IPs (via DHCP), and just renumber them when there is a need to redo their network. They just prefer not to do it this way.

Reclaiming all those class As would really only delay IPv4 exhaustion by a few months. That was true 10 years ago. That was true 20 years ago. That will be true 30 years from now, when we're all on IPv8.

It wasn't reclaiming class A space that got us 10 years, but rather NAT and the existing pool lasting as long as it did.

Of course, the only purpose of the APAP in this medication is to kill the patient if they dare to abuse it.

There is no reason that the two couldn't be unbundled so that the doctor and patient could work together to ensure that just the right amount of each is administered (likely zero on the APAP front, but the doctor can prescribe it if appropriate). This is just prohibition-style thinking, just like sticking strychnine in industrial alcohol.

There are a lot of things that can go wrong when manufacturing a pill beyond what goes wrong when you administer a pill. When you make an allergy pill after having made a super-strong heart medication, will the recipient of the allergy pill get their heart slowed down as an added bonus due to cross-contamination?

Also, there are also a lot more storage restrictions on the ingredients that go into pills than on the pills themselves. The pills contain antioxidants, preservatives, antimicrobials, and so on, and are tested as formulated to withstand variations in storage conditions similar to what a typical consumer would subject them to. The raw materials are often expected to be stored in very controlled and ideal conditions (not exposed to air, protected from light, kept at a specific temperature (which might be -70C), kept at a specific humidity (not too dry, not too wet), and so on. Since those raw materials cost millions to make at manufacturing scale anyway, the cost to store or even ship them at optimal conditions isn't a huge expense.

Then when manufacturing equipment is done being used and is cleaned out there is often analytical testing done to demonstrate that no residue remains on the equipment. This is done EVERY time the equipment is used. The tablets that are manufactured are also sampled and tested to ensure proper quality. Some of those tests can take a month to perform (like sterility testing - filter the IV fluid through a filter, then put the filter in growth media, and give anything that was caught by the filter plenty of time to grow).

I'm not saying these are insurmountable obstacles. However, printing tablets is unlikely to be just like loading up your inject printer with a few cartridges and hitting print. Compounding pharmacies today are being scrutinized very closely because there have been some high-profile lapses in quality.