russss

Pretty much all electricity markets worldwide set the unit price based on the the cost of the "marginal" (most expensive) generator running during each time period. It's a weirdly common misconception that the UK is unique in doing this.

If you paid generators what they bid, then they're incentivised to manipulate their bids to try and make the most money, distorting the market.

Almost all the wind farms and many solar farms in the UK operate under the "contract for difference" system, where they're guaranteed a fixed price per unit and have to pay back any income above that. So a lot of the money paid is clawed back through that method.

The reason the UK's electricity has been expensive over the last few years comes down to:

  - Shutdown of several nuclear plants without any replacement
  - Shutdown of coal plants and replacement with gas
  - The Ukraine war affecting gas prices
  - Clean energy surcharges on bills (which hit electricity bills a lot harder than gas bills, regardless of how clean the electricity is...)

There will be a bunch more renewables coming online soon which will hopefully start crowding out gas and driving the price down more regularly, so hopefully prices will start dropping faster soon.

A CIO is more onerous than a CLG because a CIO is a charity, and comes with the charitable reporting requirements (and also the tax benefits).

A pure CLG doesn't have the charitable tax benefits but it's still a non-profit governance structure, and you only have to file with Companies House and HMRC. The CLG is the lowest-hassle non-profit structure.

(Confusingly, CLGs can also be charities - because charitable status can act as a "wrapper" around other corporate structures - and indeed a lot of charities are CLGs because the CIO is a more recent invention.)

I wrote about this, ages ago, I think it's mostly still correct:

https://russ.garrett.co.uk/2009/10/25/starting-a-non-profit-...

> It seems to me like this is a shortcoming of the English (and American) legal system, where non-profit organizations and such can only be founded as "regular" companies and then tax-exempted. This means they are still subject to all the normal procedures surrounding companies like sales, and still tend to have one or a few owners who are solely in charge of everything.

This isn't true in the UK. There's a structure called a Company Limited by Guarantee, which has no shareholders so it can't be sold. It's an association.

These are quite common, and in fact Freenode Limited is one. It was never "sold" in any conventional sense. There is no owner of Freenode Limited, but Andrew Lee is currently the only voting member so he has full control.

I don't think there are any UK health insurance companies which allow you to claim for NHS doctor's appointments.

Presumably if you have a health insurance plan which covers private doctor's appointments, they get a copy of the doctor's notes anyway.

London, UK - Smarkets

Smarkets is disrupting the global betting industry by offering a modern betting exchange with significantly lower transaction fees than the competition. We're a well-funded company with a small, agile development team.

Our office near Old Street in London is shared with other exciting startups and has a pool table, foosball table, and fully stocked fridge.

We're heavily driven by user-focused design and a focus on technology and engineering as a first class discipline.

We write our software in Python and Erlang, and rely heavily on asynchronous programming techniques and REST. We build on a modern, open-source software stack which includes Linux, Vagrant, Flask, Eventlet, PostgreSQL, RabbitMQ, ElasticSearch, Graphite, Chef, and Git.

We make extensive use of version control, configuration management, and automated testing, which lets us deploy code to production several times a day.

https://smarkets.com/about/jobs/engineering

It's probably because modern tweet IDs are larger than a 32-bit integer. Presumably some JSON parsers aren't too hot on parsing bigints, so they give you the option of having a string instead.

The liquid nitrogen is just used for the ground tests. On the Skylon itself, this cooling capacity will come from boiling the liquid hydrogen fuel.

It's European Circuits: http://www.european-circuits.co.uk/

We really should have mentioned them in the blog post as they're sponsoring us. It's obviously a big plus to have these being fabricated in the UK instead of China due to our really tight lead times.

They are starting with engines. The rest of their ideas are simply concepts at this stage. It's a small company.

The problem is that plenty of people have tried and failed to produce a successful air-breathing rocket engine. Nobody wants to invest any money in Reaction Engines until they can prove that the core concept that the company is built on works. And proving this costs a certain amount of cash. Some of this is being financed by the ESA, in very much the same way that NASA is financing new commercial spaceflight efforts.

In the next year or two we're either going to see them succeed and garner lots of investment, or fail and fade into obscurity.

It's not a conventional jet engine - it has an intake turbine, but it's powered by the same helium cycle used by the precooler, and not by combustion byproducts.

The compressed air (or liquid oxygen when in rocket mode) and hydrogen is then combusted in a way similar to a conventional rocket engine. REL calls it an "air-breathing rocket engine", but it shares more in common with a ramjet than a conventional turbojet.

> Here incomming air is highly compressed -- and thus heated up -- due to impacting plane parts at hypersonic velocities.

The problem is that at high mach numbers, the incoming air is too hot to successfully combust. They have to chill it (using the cryogenic fuel), and chilling it would usually result in icing.

The whole point of these tests is to prove that their precooler system resists icing.

Well, this is their secret sauce, which they claim will make the SABRE engine work where other scramjets have failed.

The tests they're currently running are of the pre-coolers, which have a proprietary design to prevent icing. If it works, they will have overcome the biggest hurdle to building a hybrid scramjet/rocket engine.

Most of this is code is available in machine-readable form in the VirtualAGC Google Code repository:

http://code.google.com/p/virtualagc/source/browse/#svn%2Ftru...

Right. It's technically correct but it's misleading.

I don't think NASA holds shares in SpaceX - the word "invested" is wrong here.

I think what the article meant to say is NASA has awarded SpaceX $381m in contracts and development awards (through the COTS and CCDev programs).

> How do SpaceX communicate with the Dragon in orbit? Have they built their own global tracking/control network?

No, they're using the US government's TDRSS system (http://en.wikipedia.org/wiki/Tracking_and_Data_Relay_Satelli...), and I think they're also getting support from NASA's ground stations.

> Also, is sharing of ground facilities (fueling, launch, recovery) with other operators necessary for the business to be economic?

I don't think so - it seems like the cost of running the ground-based infrastructure is quite minimal compared to the actual operational costs of launching. Although SpaceX is using some of the original NASA/USAF infrastructure at Cape Canaveral and Vandenberg, it seems like they've built most of their facilities from scratch.

We used MarkMonitor at Last.fm. It's probably best to think of them as an abstraction layer above the registrar level which handles any residual pain that might generate:

* They will register all TLDs. A lot of more obscure ccTLDs (e.g. Norway, Hong Kong, Argentina) require domains are registered with a local company. MarkMonitor will set the local presence up for you and charge you a fixed fee. Some other registrars support some of these TLDs but none support all of them, so otherwise you're split between registrars. If (for whatever reason) MarkMonitor can't register a domain themselves, they'll go as far as getting a credit card out and using another registrar who can.

* They invoice you: you don't have to worry about expired credit cards or any other payment problems. You get one invoice every month for all your domains. (If you've ever had to deal with submitting expenses receipts for 80 domain renewals, this is definitely a bonus. Now consider that most huge companies have tens of thousands of domains.)

* They handle the domain "sunrise" applications for you (this is apparently a big brand protection deal if you're a big company), so you'll regularly get emails saying "do you want to register the .xxx domains for all your brands?".

* They do SSL certificates as well, and they automatically get renewed and invoiced.

* You get an actual friendly human account manager to talk to.

I can't remember what their actual pricing is. The price per domain is perhaps 2-10x more than other registrars. I'm not sure if they charge a recurring fee.

The reason for that is there's a an important satellite ground station on Svalbard - it's in an ideal place to communicate with polar-orbiting satellites:

http://www.ksat.no/node/77

http://en.wikipedia.org/wiki/Svalbard_Undersea_Cable_System

As mentioned in that man page, the /etc/lsb-release file only contains details on what version of the LSB the distribution conforms to (if the distribution isn't LSB-compliant, the file shouldn't exist). It doesn't have the version of the distribution itself, which lsb_release finds in distribution-specific locations.

Yes. At Last.fm, Germany was the first country we had to censor our content in.

If you try and view content which is deemed inappropriate by the German government (fun fact: even the list of inappropriate content is restricted), you get something which looks like this:

http://www.lastfm.de/music/Landser

I believe these pages are still available if you try and visit the English last.fm site from Germany.

WHOIS is generally very reliable - there are identity checks when you register for IP space (especially with ARIN, which is who provides Google's IP space).

One other way you can check is by looking at the organisation handle in the WHOIS - it's GOGL, which is the same org handle as you'll find on all their IP addresses if you WHOIS them. That ties up with the ARIN site:

http://whois.arin.net/rest/org/GOGL

I've done a fair bit of network engineering and I'm convinced that this is actually a Google address.

GKBO appears to be a genuine Google initiative - its domain is registered to Google and served from their nameservers: http://www.kbo.co.ke/

That Sprint problem was caused by Sprint "borrowing" DoD addresses to use in their internal NATed mobile network. This wouldn't be an issue with a standard web site like Mocality which is directly connected to the Internet, instead of via NAT.

I trust Mocality's technical chops enough to believe that the IP traffic is coming from where they say it's coming from.

As additional proof, the callers claim to be from Google, and Getting Kenyan Businesses Online is a genuine Google initiative. So a lot of things wouldn't add up if it turned out not to be Google behind it.

People do use hydrogen, but most balloons still use helium because it does provide a slight safety advantage when launching. (It's also sometimes an insurance requirement.)

I expect to see an increase in the number of hydrogen-filled balloons as helium prices rise in the next few years.

This is known as a "Duress Code" and it's sometimes found in alarm systems as a code which will disarm the alarm and still call the police.

http://en.wikipedia.org/wiki/Duress_code

To elaborate, this manual refers to the Soyuz rocket, not the spacecraft which is flown to the ISS. Specifically, it's the interface documentation for attaching your payload to it. (The Russians/Soviets have a long history of naming their rockets and spacecraft identically.)

The ESA are collaborating with the Russians to build a Soyuz launch site at Kourou in Guiana. It's pretty much a complete clone of the site at Baikonur:

http://1.bp.blogspot.com/-SBynwXjyxKI/TZiECQqNtGI/AAAAAAAACS...

This gives commercial customers an increase in payload capacity compared to launching from Baikonur because Kourou is nearer the equator (especially relevant for geosynchronous orbits). Plus you get the legendary Soyuz reliability.

I believe they're planning on the first launch from Kourou before the end of 2011.

"This post is largely taken from a proposed redesign I sent out last year."

Yep (I'm organising it!)

The issue at Fukushima isn't criticality, though. All the reactors were safely shut down seconds after the earthquake hit, and no new fission has occurred since then.

The problem is that even a shut-down reactor core generates significant heat due to the decay of fission by-products, and that's what caused the problems.

That said, there are other reactor designs which are more inherently safe.

My bad - it's been a while since I touched Rails and my memory was rusty.