Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?
Slashdot Deals: Deal of the Day - Pay What You Want for the Learn to Code Bundle, includes AngularJS, Python, HTML5, Ruby, and more. ×

Comment Put it in the library (Score 1) 25

NO language needs a garbage collector, though an option to use one selectively would be nice.

Let's split the difference and say a language needs a garbage collector in its standard library that a programmer can choose to enable. C++ calls its reference-counting garbage collector std::shared_ptr.

Comment Re:Cost of access is key. (Score 1) 309

That was not my point. Ofc we can improve ISP. No idea how much that improves either 'performance' or drops price.

It improves performance a *lot*. As for price, it depends on how expensive that rocket system is. For first stages, an improvement in ISP's effect on the size of the rocket isn't that much greater than linear. But the further up the delta-V chain the engine is used, the more of an impact it has on everything that was used to get it there. An extra hundred sec ISP on a first stage might reduce the system mass by a third; on a second stage up to LEO, maybe cut it in half; on a kick stage for a Mars transfer orbit, maybe cut it by two thirds. On an ascent stage from the surface of Mars... well you get the idea. Shrinking down a rocket to a small fraction of its size - fuel, tankage, and engines - well, that's really significant. ISP is very, very important for upper stages. So you can afford to pay quite a bit for those top stages if it improves their performance. Just not an "unlimited" amount.

There is no way a high tech electrical engine will improve its performance by 10% regardless how much money or time you put into it: the efficiency is already between 98.5% - 99.5%, up to 99.9% in some cases.

This is getting a bit offtopic, but at least the electric engines in EVs don't usually run at nearly that high. Depending on the type they might average 85 to 94% on average. It varies over their load cycle.

Regarding rockets: there is simply not much margin anymore in changing the form of the exhaust tube, burn chamber etc

Actually you can. The general principles of how rocket engines work are fixed, of course - your exhaust will never exceed its local speed of sound in the throat, and then you want to expand it as close to ambient pressure as you want. But the details vary greatly. There's bell nozzles, linear nozzles, annular nozzles, aerospikes, throatless nozzles, atmospheric wake compression, and on and on. There's tons of different ways - developed, in development, and in theory - to pump and inject your propellants - where they need to be pumped at all. Even many propellants that are traditionally thought of as being in one state can be implemented in other states. There's various ways - developed, in development, and in theory - to prevent nozzle erosion. To improve regeneration. To reduce mass. And on and on and on. Rocket combustion is a rather complex thing and we're still trying to get a handle on it. Do you know that we still really don't know how aluminum burns in solid rocket propellant? There's something like five different competing theories. I mean, things like this are a Big Freaking Deal(TM), especially when such small improvements in upper stage ISP have such significance for lower stage mass. And even on your lower stages there's a lot of things that have a big effect on your system cost. For example, how to stop resonant shocks from ripping them up - a lot of people don't realize that one of the main benefits of adding aluminum first stage to propellant mixes is that the droplets of burning aluminum damp shocks. (yeah, it increases ISP too by raising the exhaust temperature, but it also has disadvantages, such as not contributing to expansion, slowing down gases (particularly near the nozzle), and impacting/eroding the throat (or even forming an accumulating slag)

Re, nuclear+chemical. There are proposals for this. The main issue isn't efficiency - the extra chemical energy doesn't make that much of a difference - but thrust. The downside to nuclear thermal is that the reactor is so heavy (fission is like that, unfortunately) that the mass ratio is only something like 3-4:1. That's really bad (you generally get 15-20:1 or even better for a chemical first stage). So the approach is to inject oxygen early in the ascent phase for added thrust, but only run on hydrogen higher up when gravity losses are lower. I'm really not that sanguine about nuclear thermal rockets getting a serious development program any time soon, though. The public overestimates the risk, of course - not only am I sure they'd well seal the fuel elements against whatever damage would be incurred by explosion or reentry, but there's the simple fact that the fuel is "fresh", not contaminated with the more hazardous actinides. But it's going to be a hard sell. And a really hard development project, if they ever did try again. Gigawatt-scale flying nuclear reactors that pose radiation hazards during assembly and test aren't exactly childs' play.

Comment Re:The guy aint no Sagan... (Score 1) 309

You forgot to exclude operational expenses.

Yes, people to run robots and comm time on the DSN. We're not talking about massive expenses here. The real expenses are the capital costs.

And also didn't mention that you can't just lob chunks of metal straight to Earth's surface,

Actually, you really just can. Even random rocks from space - not shaped for optimal entry shape, not cemented together by anything yet what nature chose to gie them - do this all the time. They have to be between a certain size range (too little and the whole thing ablates; too large and it explodes, either in the atmosphere or on impact), but the random creations of nature do it; delberately shaped and sintered projectiles should have no trouble with it, with (proportional to their mass) relatively little burnoff.

You would, of course, need a rather large area designated as the impact area; even with very precise aiming, by the time they get to Earth and undergo reentry the random variables will spread them out over a sizeable chunk of land. A large salar might be ideal, since they get resurfaced periodically so the impacts wouldn't be damaging the landscape.

By your same logic, the mining of minerals on Earth would be zero dollars per gram if the equipment was solar powered and automated

It's almost as if I didn't discuss capital and ongoing costs in my above post.

Launch costs really are key to the rate of development at the very least, in that they limit the rate in which funding can be raised for the necessary exploratory and test craft to be launched. Even if the economics for operating a mine on a NEO works out really well at present launch costs, you have to prove that you can do it before you can raise the billions to build it. And to prove that you can do it you have to launch a number of missions while you're still relatively poorly funded. They face the same problem that Bigelow has faced - a probably reasonable business plan but the early phases hinging around factors that they don't control.

It does nobody any good to pretend that the lack of a space economy is because investors are cowards and morons

I think you need to go back and read my last post again, particularly all of the "it's too early to say"/""we don't know"/"but time will tell"/etc lines. I'm not saying that at all. I'm saying that there very well could be a compelling case for asteroid mining even without any radical changes in space technologies. But there's a great deal of work to prove that before we can get to that point.

Comment Even if the sky is falling down (Score 1) 102

well... when the sky falls, then chickenlittle should worry

In this model, I've spotted the pieces of camouflage that the aliens are assembling.

my sources in the security community get those new adserver names as the pop up easily too.

Not if they're like a328bc97.someadnetwork.example. Even eight hex digits would require four billion lines in a hosts file.

Comment Re:Not meant to be a good device but to undercut C (Score 1) 188

Two of the pins (+5V and any GND) on the 40-pin connector can be used to supply power instead of going through the USB port. That's what I did with my beer-fridge controller: power for the whole system comes through the barrel connector on the 1-Wire/I2C interface board in the middle of the stack.

Guess I shoulda looked at the pinout before leaving that comment. What do you need for I2C? Is it more than some resistors? Hmm, I looked and it seems you just connect up the pins. Internal pullups? on-board? Leaves it to external? I am way too lazy to hunt through the docs to find out. Did you put in some fuses or something? My experience with I2C is so far limited to connecting Arduinos to IMUs and so on. Also did the SPI sdcard thing there with the sdfat lib. Hooray for electronic tinkertoys.

Chemist who falls in acid will be tripping for weeks.