ballooney

Well, I'm sure you wouldn't really believe for a second that this kind of surveillance is not also happening in the US, if GCHQ are also doing it, so go on then, challenge it and see how feasible it really is.

No, it’s different people doing it, but the same few people funding it. Like VC.

He hasn’t done that with rockets. He wants to eventually, but he hasn’t yet. Let’s not let hype cloud facts.

I only really see it used on potential employees, rather than potential customers. It sounds better than the more-correct 'come and sacrifice your best years on this pointless app that might make me rich but probably will crash and burn like all the others'.

This example doesn't do the entire numba project justice, but if you've ever written a for-loop in a bit of python code that does number crunching, you'll notice how much it slows everything down, and the numba jit provides a decorator that yields an extremely quick win to get often 1-2 orders of magnitude of improvement in calculation time. It's less of a win if you're only every working with already vectorised data structures and algorithms.

I was confused by your comment though, specifically the idea that you are using tensorflow because you 'mostly work with simple data that doesn't require complicated calculations'. This seems very contradictory. Have I misunderstood?

    > he created an awful lot of value for them

Has he though? No one who bought Uber for $70bn right now would ever get their money back, despite that being it's 'value' (I know, I must be new here). This is all just fantastical and silly, a constant game of persuading investors to play along. Unless there is some great master plan for pivoting into a business that's actually feasible as a business, he's no more a great CEO than the Wizard of Oz or Bernie Madoff.

I've noticed (as a non american) that every single book I see in american airport book shops (at least the sort of business/self-help ones which seem to constitute 95% of american airport book shops) follow this exact title style: $PITHY_PHRASE: $MEANDERING_SUBTITLE_THAT_IS_A_BIT_TOO_LONG

It's very unimaginative.

This is correct, the power density of electric motors (and the batteries to power them) is still very far short of what you can get from a well-designed turbine-based pump (that is, a turbopump), but there's so much ancillary plumbing and and stuff for a turbopump (like the precombustion chamber that has to generate some hot gas to drive the turbine), that doesn't scale down all that well to the small scale of the engines on the Electron, that the power density advantages of a turbopump tend to asymptotically tail off. Also, they require a lot of expertise to design, compared to almost any other kind of pump. There are turbopumps and turbopumps, though, and here I will paste in an HN comment of mine from a number of years ago:

""" There are three kinds of rocket engine cycle (well, there are maybe more but these are the three that have been flown historically). The Expander Cycle, the Staged Combustion Cycle, and the Gas Generator cycle. I'll mention the last two.

Merlin, as the article mentions, is an example of a Gas Generator cycle. In this cycle, you take off a little bit of fuel and oxidiser to burn outside the main combustion chamber, to generate some hot energetic gases that you can exhaust over a turbine. This spins the turbine up, which is connected to a shaft with a compressor on the other end. The compressor increases the pressure of the propellents so that they can be injected into the main combustion chamber. This assembly (turbine, shaft, compressor) is called the turbopump. It's necessary because the engines require very high flow rates to get the thrust they need, and that has to be at a high pressure - higher than the pressure of the combusting gases inside the combustion chamber, else you wouldn't be able to inject it!

Back to the bleed-off to drive the turbine. You usually don't want a perfect stoichiometric mix of fuel and oxidiser for this, or even close, because it generates extraordinary hot gases that no turbine would last long in (The turbines are spinning at many tens of thousands of RPM usually so would be subject to much higher forces than the actively cooled walls of the main combustion chamber). For this reason you usually have a large imbalance of one propellent to the other to keep the temperature down. Usually you run with excess fuel, or 'fuel-rich', as the opposite - oxidiser rich - means you have hot oxidising gases which are harder on the metallurgy. I do know of some russian exceptions to this, though, where fuel rich would have left sooty deposits in the plumbing (The materials science employed in the turbines was apparently so witchcraft that when the US got intelligence of oxidiser-rich turbine precombustors, they thought is was deliberate counterintelligence from the russians to get them to waste billions researching the impossible). The gas generator cycle, as the article mentions, dumps this turbine exhaust overboard separately. The problem with this is that there's a load of uncombusted fuel in this exhaust, which you're just wasting, and this hits your rocket performance - the Specific Impulse ( I_{sp} ), as you're not getting as much bang out of a given mass of fuel as you could.

The answer to this is the Staged Combustion Cycle, where you also inject the exhaust of the turbine into the combustion chamber to finish off combustion. The performance of these engines is higher but the thermodynamic balance to design a working system is a greater challenge, and some of the engineering is a bit harder too. Staged Combustion engines are mostly russian, although the Space Shuttle Main Engines are a US-design example of Staged combustion. """

Staged combustion engines are extremely efficient and on a big engine no electric pump system will even touch them, unless there is some materials-science breakthrough that will allow us one or two orders of magnitude improvement in flex density in electromagnetic materials. Electric pumps will probably remain in their niche for small engines and satellites.

I do think that small turbopumps are worth further research, although I don't know if the market needs higher performance small engines over more cheaper-to-produce small engines, but certainly there was fascinating work done in the uk in the 70s with tiny turbopumps (about the size of a coke can) that ran at hundreds of thousands of rpm, with a power of megawatts, and compressors very cleverly shaped to run sustainably far beyond the cavitation point of the fluids, which is usually the point at which you can't pump anymore, in traditional pump design literature. In combination with an expander cycle you could probably produce some extremely high performance, simple, small rocket engines. Maybe.

We live in exciting times.

I've done space projects with US companies and so am fully marinated in ITAR processes. In answer to your question: No, you can get around ITAR by just being careful about what information is imported and exported to and from the US. If there was a US engineering office and an NZ engineering office and all the engineering effort was evenly distributed among the two, it would be a nightmare as you'd be 'exporting US technology' constantly, but if all the important stuff is being done in NZ and very little know-how has to go back-and forth across the virtual border with the US then it's more manageable (still involves lawyers and experts though). It doesn't matter than the US owns it financially, it's the transfer of knowledge and stuff across the border (virtual and real) that is the hard bit. As this is basically a NZ company which had to do a paperwork fudge in order to accept US investment, ITAR is not such a problem.

You might as well just send up a second one, rather than send something that is 95% of the first one plus a very complicated system of robot arms to do arbitrary repairs and replacements.

This is surely some markov-chain HN parody account. Do I win anything if I'm right?

If, as suggested by the article, it's acrively telling you you're somewhere else, that's an order of magnitude more sophisticated than the alibaba jammers which truckers use to just stop their trackers.

Not really, there are commercial agreements in place between both SpaceX and Boeing, and Nasa, as part of its Commercial Crew Programme. The current estimates are that they will not launch until 2018, three years behind the original schedule:

http://www.parabolicarc.com/2016/09/01/nasa-oig-report-delay...

I also have customers who have been sold slots for their satellites on Falcon Heavy which is now 4 years behind schedule (the first demo flight was originally meant to be 2012, with the first commercial flight in 2013, as yet nothing has flown, with the current estimate being 2017). It's frustrating for these customers.

I point this out here just because HN can have too many credulous website designers saying [to the effect of] 'No but Elon will be on Mars in 10 years so we should just fund private companies' and the reality for everyone in the industry is somewhat different. The Indian Government is currently about the cheapest and most reliable commercial satellite launcher in the world with its PSLV vehicle:

https://en.wikipedia.org/wiki/Polar_Satellite_Launch_Vehicle

No one does.

That seems like the readonable answer on the part of the barista.

I don't mean to sound snarky in the least, genuinely curious, but haven't a large proportion of tinkerers, especially with an EE bent, tried to put a simple lisp interpreter on some limited hardware at some point in their lives? I certainly know that a bunch of people learning FPGA programming in a user group i once belonged to all tried to put a lisp machine on one after working through the last chapter of SICP, and there's that lovely article about lisp interpreters running on old JPL probes back in the day. What about this makes it revolutionary?

That's a good thing. Too much of the internet is people with nothing to say, saying too much.

Spot the american! (I mean that in a good-natured way). There are a few 1700s restaurants in london, plenty of 1800s, and of course if you look at pubs in the UK then 1840 is practically modern. One pf my local pubs dates to 1381. The food is probably better now than it was then.

A very HN correction here but 'space shuttle' refers to the whole stack (orbiter and fel tank and SRBs) and what you seem to be referring to as the 'space shuttle' is actually called the orbiter, a component of the space shuttle. But everyone gets it wrong. Mobile makes it difficult to add links but its explained on the wikipedia article for the space shuttle orbiter, a distinct article from the space shuttle.

This is a very interesting talk about writing the software for the Curiosity Mars mission by the head of Software Reliability at JPL, Gerard Holzmann, who also wrote these 10 rules:

https://vimeo.com/84991949

It will officially take two years to leave once the process has started, which it has not been yet (that's what the Prime Minister declares). Whether or not you subsequently have to pay import taxes after Britain has left is also something that will have to be negotiated.

I am a physicist by background and rocket propulsion engineer by career. I am pretty handy at writing signal processing and control stuff on microcontrollers and FPGAs, and writing physics simulation code in C++. I thought, because I am therefore Really Smart (tm), I must be able to knock up a webpage because how hard can it be? I basically died a death within a week, surrounded by sharks in a pool of confusion about javascript and frameworks and doms and salt and pepper hashbrowns and bcrypts.

Reciprocally, I have visited a few hackspaces and witnessed vignettes along the lines of someone saying 'I know Javascript, therefore I'm an engineer!' before they then cut their hand open in a completely predictable way with an angle grinder, having just rebuked you for trying to 'interfere' with their use of said angle grinder because you could see the wheel nut wasn't tightened down properly (this is a true story).

The moral is: don't assume other's backgrounds, and don't over-assume your own. We all come at this stuff from different directions and a pointer can help some of the audience, even if its redundant for others.

I'd expect it to be up to 100 celcius at sea level pressure, because it had just been boiled. If it's hotter than that it wouldn't be a liquid. How can less-than-boiling be 'well beyond' the expected temperature of just-made tea or coffee? Your accusation is totally absurd.

Maybe I'm just being too European.

His lectures on inference were the best and most worthwhile course I took as an undergraduate, and his text book is good enough that I have two copies at home and one at work. He had a big effect on how I look at the universe. He later became one of the only voices on energy policy that I trusted, and enjoyed hearing his clear thinking on the national news. Thank you, djcm. The world owes you quite a lot and probably has yet to realise it.

Elon Musk could start an energy company which renders down baby seals and burns the blubber and HN would say its usual thing like 'Actually this is a really smart move by Elon' (everyone here seems to be on first-name terms with him). This place would be greatly improved by some perspective.

He's built a very cool rocket company, but on HN people take the CTO title at face value and assume he's a genius rocket scientist. He's built a cool car company, and is right to focus on battery technology and solar being better than biofuels etc etc. I admire him a lot, but i think a lot of people project their ideals onto him as a sort of symbol of the primacy of the Valley Way, which they're staking their twenties and thirties on. Whilst taxonomy doesn't really achieve much, I sort of think of him as more Henry Ford than some sort of Tony Stark/Einstein hybrid that is often seen here.

Like I said, there's a lot to admire, but he's just a guy still. Petty, tantrums, lots of SpaceXers i know (current and former) talk about how really you have to go to Gwynne Shotwell for some sober leadership and patience when he jumps the shark a bit, and she can speak truth unto power and figure out a way forward. She never gets mentioned from from what one hears in this industry (I'm in it to) she deserves a lot of credit.

I interpreted it as:

Elegant, clean design (certainly compared to the old 8-bit PICs). Good taste. FOSS-friendly.

Microchip might be a Texan oilman in this comparison. Just keep pumping them PICs out don't listen to those commies with their 'GCC' and their 'lee-nux' - sounds foreign and suspicious.

This is obviously a tongue-in-cheek characterisation but I think that's what the OP was getting at.

nbconvert infact does use pandoc for most of its internal mappings between markup languages. See the first 'Note' box:

https://ipython.org/ipython-doc/1/interactive/nbconvert.html

I work in rocket propulsion research and have worked with, and around, both of these propellants, and I think you exaggerate to the point of being incorrect. You can smell hydrazine long before it's dangerous - it smells a bit fishy, somewhat like ammonia. It has to get really quite bad before your body-mounted gas alarms tell you that you're above recommended limits.

I'm more worried by NO2 but as a fume it's bright orange so one can normally see a cloud of it coming and make plans to head somewhere not downwind. If you do get caught by some, protocol is to lie with your head at the bottom of an inclined plane for several hours, so the the fluid can drain out of your lungs (and you call an ambulance of course).

When propulsion tests are happening on site we keep emergency breathing apparatus under our desks incase there's a containment failure and we need to get out. Usually though we just test when the wind direction would take anything away from the office/workshop.

I'm not bothered about kerosene but I have a healthy respect for LOX; it's as easy to kill yourself with it as with hydrazine. Where breathing apparatus and chem suits will keep you safe from hydrazine pretty much, LOX doesn't forgive any dirtiness, grease, human hair, or the wrong choice of o-ring material, especially when pumped at high speeds and pressures. Different risks but LOX is arguably more likely to bite you.

Here's an often-cited LOX safety video of the sort you probably couldn't make today:

https://www.youtube.com/watch?v=QioE_OX4vP8

I recently had to sit through a presentation by an X-er in which was played a very smug GoogleX video that spliced an annoying amount of apollo- and shuttle-era space exploration footage to pump up what was essentially floating cell towers (Loon). I found it quite disrespectful to the apollo engineers, and grasping of Google PR, to try and equate what they're doing to landing on the moon. It went quite beyond the bounds of what is reasonable with a straight face, for my tastes.

The arduino is really just a breakout board for a microcontroller made by Atmel, with a simplified development environment to lower the barrier to entry. The microcontroller chip itself has always been a couple of dollars, and was before the arduino was invented.

Really you could say that paying $20/board to break out a $2 chip was the insane bit (of course, if it makes the difference between you tinkering and not tinkering, totally worth it). It certainly hasn't been the cost of making arduinos that's been holding back your 'amazingly disruptive disposable electronics' - no chinese factory will have been paying more than a few cents in total parts costs for their mass-produced widgets of this sort for years now.

If you want garage-tinkering-friendly microcontrollers, can I tempt you with the newish NXP LPC810 [1]? It has a 32-bit core which can run at 30MHz (as compared to the 8-bit core running at about 16MHz on the arduino, iirc), and they're $1.37 in single units, less than 60 cents if you want 1000. ARM maintain a whole GNU toolchain [2] to develop on them - it couldn't be easier.

[1] http://www.digikey.com/product-detail/en/LPC810M021FN8FP/568...

[2] https://launchpad.net/gcc-arm-embedded