Be careful how you change the laws. Copenhagen Suborbitals has the sea launch platform Sputnik. That exists exactly because small ships without passengers face little regulation. Obviously changing US laws will not cause trouble for Sputnik, so in that way the example is contrived, but it would likely have been quite a hassle for Copenhagen Suborbitals to get the vessel approved. It is not exactly a typical ship, so the paperwork could end up quite substantial.

I have every reason to believe that Sputnik is safe to use, and the various crew seem to be safety-conscious to the point of being borderline paranoid. It would be a shame if red tape stopped a similar group of people in the US from making something spectacular.

Traditional automatics are just as obsolete, of course. When a semi-automatic can shift gear in 50ms, why would you want anything else?

We are talking Formula 1 here. Supposedly the pinnacle of racing, the ultimate series. Why should they be saddled with manual gearboxes? Performance cars do not have them anymore. There are rally and stock car racing series for those who like that kind of thing.

There is hardly any auto-anything allowed in Formula 1. Manual gearboxes are obsolete technology, it would border on the ridiculous to use those.

This would perhaps be slightly relevant if say 0.1% of the processing was handling transactions. Bitcoin mining is about useless hashing, not about transaction processing.

The only actual value provided when a new bitcoin is found is a bit of joy for those who like numbers that hash to certain types of values.

People get in hot air balloons all the time, despite their rather dismal safety record. Hot air is not viable for airships.

There are currently three possible lift gases for airships: hydrogen, helium, and water vapour. Hydrogen is out for safety reasons, helium will be too expensive, and steam is difficult because the airship has to be really large to avoid too much heat loss.

Good luck convincing anyone that hydrogen airships are safe.

It would be preferable to use something that reacts less easily with oxygen than hydrogen, in many cases.

Fair enough, we can teach all X applications to handle disconnections from the server and somehow wait for a signal (coming from where?) to create a new connection. Good luck with that.

Meanwhile, every other remote GUI has had this problem solved from the beginning.

What is the use case? When I want remote display, I want something which does not kill all my applications when the network connection drops. That is just not useful behaviour.

Forget it. Anything vaguely modern renders client-side and gets it wrong.

X applications die with the network connection -- they cannot survive when the machine running the X server changes IP or hibernates. They are tied to one X server, so you cannot move them from your laptop to your tablet.

It has been at least 10 years since I used X forwarding for anything except the rare GUI installer or similar short-running application. VNC is much more useful.

Itanium killed the Alpha (which was admittedly dying already), PA-RISC, MIPS for servers, and almost took out SPARC. Lots of the development was paid for by HP. The only high-end Unix CPU architecture left is POWER.

I am pretty sure that Intel did not intend for Itanium to fail in the market, but in retrospect the outcome for Intel has been close to perfect. The only challenge for Intel right now is ARM, and a successful Itanium would be a hindrance rather than a help when fighting ARM.

That is how CISC chips work, indeed that is what makes them CISC. The whole point of RISC is to not do this translation, so it makes no sense to say that "x86 machines are RISC". x86 will be RISC the day the underlying instructions are exposed to the programmer, but that day is far off because it would mean reducing performance rather than increasing it.

Why? T1 routers these days are legacy devices with very few units sold. Of the few units shipped, most are likely low-performance devices like the 1900. If you buy something non-mainstream, it is usually more expensive than a mass-market item.

You are completely right about the ASR 901 though. It would have been a much better choice than the ISR 3945. However, the routers were purchased in 2010 and it seems the ASR 901 was not announced until 2011.

My guess is that they were integrating all the old lines into one network along with the new ones, so someone would have had to go to each site to make that happen anyway -- and installing a new router could easily be required for that. But yes, perhaps those sites could have waited a bit longer before joining the new shiny future. Maybe the contract with the old provider had ended or something silly like that.

The problem is that the many of those sites were getting fiber out there. The state wanted a single device that could handle both the legacy T1's and the new fiber connections. Cisco really ought to have told them to go with whatever their cheapest T1 model is these days and then replace the router when the fiber is actually installed. Cisco is certainly to blame for not doing anything to help out.

However, the state is certainly to blame for not letting someone with a little bit of experience take a look at the bid.