Strategic thinking as an exercise

Thursday, 13 February, Year 6 d.Tr. | Author: Mircea Popescu

It all started with

asciilifeform mircea_popescu: from the Fun With Widgets dept.: cryptome.org/nic-ssh-rootkit

Yet another article describing what we've known for a while : secure computing is in fact rendered impossible by the hardware itself, what with DMAi, what with defective separation of controlii with CPUs embedded in all the peripherals anyway and so on and so forth.

Basically computer hardware is the house that Jack built, with everything including bypasses for everything else because everyone fucks up everything all the time and nobody can be trusted to maintain a spec or even specify anything in the first placeiii.

Basically computing has gone through a revolutionary phase of very limited means, Valleyforge in the 40s, to a plentiful phase of happy abundance - the 70s in the 70s - and has since then simply been piling debt and insanity upon debt and insanity. Incredible how closely the silicon (which is and has for a while been a purely east Asian concern) follows the follies of the Western world. If you noticed that the Champagne of 1795 was murky and confused, whereas the Champagne of 1799 all turned red and then in Champagne of 1803 it was all bitter for some reason... then you'd have something comparable to how the chips have gone.

Anyway : the observation leads to two lines of reasoning. The first :

    mircea_popescu

    Arrigo Triulzi: The only way would be trusted computing if implemented properly and without the DRM halo which it normally carries.

    This is not actually correct. the correct way is triangular network designs. Much cheaper and definitely practical.

The second :

    asciilifeform Those 'willing to open their third eye' will see the market for deliberately-minimal HW - e.g. NICs without general-purpose CPUs, etc. (WTF does a NIC need a dual-core MIPS?)
    mircea_popescu Indeed. Which ironically will be a major chance for the third world. It's funny how this battle goes over the years. Africa, South America and central Asia are colonised, because they suck.iv So they suck colonist cock and try their best.v However, by the time they get their tech to the level of making their own lightbulbs and other peripherals - well suddenly the ex-colonizers decide the kind of lightbulbs they can make are illegal to use.vi However, just about at the same time the most important industry of the colonizers, ie silicone processing, has "progressed" past the point of usefulness. Which opens a door just in time - a Brazilian made "ancient" technology NIC could command a premium in the marketplace. All the while CISCO reports quarters of zero sales, like it did last Q. Logically we'd expect this producer to get no likes on Facebook.

Triangular networking is something known in the practice of governance since about the days of Pericles, which is a few years before your time. Specifically, as explained in the rural vs urban dispute, technology is ideologically neutral, which is to say it doesn't help the "good" guys any more or any less than it helps the "bad" guys, for any definition of good or bad you may arbitrarily pick. Therefore, unless you're willing to abandon your ideology and embrace my neutrality, you're stuck being a Byzantine general and using byzantine troops. No big deal, Byzantium survived centuries in this manner. Sure, surviving sucks, but hey, you want that bed, you get to lie in it.

Thus : your networking consists of an exit node (which could be an old piece of crap desktop running openbsd since 1997), a sniffer, which simply reads the intranet data, and your own computer, whatever it may be. That's it : you know what should and what shouldn't be going through the network, and when. This allows you to attack any attacker into the fucking ground, just set up your sniffer to look for things you know shouldn't happen, things you know might happen in some conditions and check the conditions, things you know the attacker might need and so on. And if you suspect the sniffer may end up enlisted in the conspiracy that's okay : hire another sniffer. This is exactly, but exactly what governments do, not for any reason than because they have to. There's exactly nothing else they could possibly do, and that's how you end up with a ministre déléguée auprès du ministre du redressement productif.

The cost to you in old hardware is to the tune of 500 dollars a pop, the cost to the attacker isn't to the tune of another 500 man hours a pop, his workload increases exponentially. If you have the resources and the motive to run a large enough farm which other people have the motives to attack enough times you can even obtain experimental values for the base of that exponent, simply by keeping track of successful attacks and how deep they were. So you can actually ensure your safety strategically, which is to say by economic calculations, the only way to be actually safe.

As to the other consideration... if you read Romanian your starting point is Lovi-v-ar ecologia, bre. Including the comments and trackbacks, this thing discusses the matter exhaustively enough. And if you don't read Romanian... well I guess then you're stuck figuring it out on your own. Not like it's so very hard.

———
  1. Direct memory access, a naive, Windowesque approach to hardware design whereby peripherals - such as video cards of network cards - can directly access the RAM or even each other. []
  2. If you make a Central Processing Unit then give it the monopoly on flow control - not as a presumptive but as a factual. If you don't give it monopoly on the flow then don't call it that. []
  3. No, the idiocy of Bitcoin's power ranger team is in no way singular, they're just following what they see everywhere around. The idiocy of expecting to brew the revolution out of the methods and mechanisms of the ancien regime doesn't strike them in any way, mostly because their brains are stuck down there, several layers under such etherated strategic considerations. You wouldn't fault an ant for ignoring the clouds, after all.

    And yet, amusingly enough, actual ants have a keen if instinctual understanding of clouds, it's just these braindamaged drones the later failure of our civilisation has bred that are completely myopic and content in their myopia. []

  4. No two ways about it : they sucked, because the people living there sucked. Maybe it's because they were retarded, maybe it's because they were lazy, maybe it's because they were merely confused from listening to idiots going on about "cultural diversity" or whatever that bullcrap was at the time. Nevertheless, if you're not capable to defend yourself in the field faced with any enemy you have no culture. Period. []
  5. This incidentally is exactly why men are killed off in a conflict whereas the women are merely raped : they're much better at sucking cock, which expression is used metaphorically to describe the female superior ability of adapting herself to new living conditions and power structures. Which is why the woman leaves the house of her father to live with her new owner rather than the other way around. Obviously exceptions abound, but they are exceptions. This is the rule. []
  6. This is exactly what happened, and it was certainly not happenstance. Which is why all the pretense put up from the top all the way to the bottom about "ending poverty" and "helping poor countries" and whatnot is a despicable exercise in hypocrisy. Hey, Gates ? The third world doesn't need the poison of your ill gotten cash rewards. That hurts it and its chances, as you well know. What it needs is to be able to sell its perfectly safe and very useful incandescent lightbulbs without you idiots pretending like an ampoule full of mercury giving off a ghastly evil light is so very much better. Buncha assholes... []
Comments feed : RSS 2.0. Leave your own comment below, or send a trackback.

2 Responses

  1. Good article

  1. [...] [↩]AD didn't start at 0, you know. [↩]One Peer can certainly operate -- and indeed is encouraged to -- run multiple nodes [↩]I struggled to link directly to the PRB list item in the article, [...]

Add your cents! »
    If this is your first comment, it will wait to be approved. This usually takes a few hours. Subsequent comments are not delayed.