On pretending “Googling” is still a thing that works any more than “USG”

Over the last few days, I’ve been fiddling around with my remote bitcoin node, trying to compile the enemy’s version of the source code under the (admittedly naive) assumption that it’d be easier than the Bitcoin Foundation’s (the real one, that is) version. I have to learn to walk before I can run, right?

Well, that’s what I thought anyways. But trying to compile 0.9.3 on a Debian 7 x64 Minimal machine has been anything but straight-forward. In fact, after two days of beating my head against libdb this, autoconf that, and samba the other thing, yet being a bit stubborn about it (and admittedly naive), I finally massaged the thing the right way and it started downloading the not-inconsiderably-sized blockchain. Great success!

Or so I thought. A day later and “bitcoin-cli getinfo” was revealing that my node was building blocks really rather slowly in the 220,000 range, but it eventually overcame whatever barriers were there and continued on its merry way. Excellent, I thought, all’s well in Wellington! Yet again, not so fast. A day later still and the blockchain was now at a dead fucking stop at block 322,082 and was showing the following error message:

bitcoin-cli getinfo 0.9.3

Running “bitcoind debug” yielded the following:

cannot obtain lock on data directory

Curiously, despite saying “Shutdown : done”, the broken software did no such thing. It continued to chew up memory until I told to it actually stop. So trustworthy, this thing, wouldn’t you say? In any event, giving it a minute to catch its breath, I asked it to -reindex, which is supposed to “rebuild the chainbase data structure.” Ok, fine, it’s worth a shot. But a couple hours later and it was jammed again at block 322,082 and using 97% of its 512 MB of memory. Perhaps if I’d listened more closely listen to an astute commenter

Apparently they’ve made changes to 0.10 making it behave better; 0.9 and lower will crash permanently closer to the end of catch up process because 512Mb Ram + 512Mb swap is not enough

I might’ve saved myself all this trouble in the first place. But hey, guy’s not in the WoT, what can I tell ya. Plus, I could use the unix and command line practise.

Now this is the point in the article where I point the obligatory finger at Gavin Andresen and his merry band of wreckers.i Requiring 1 GB of RAM + 1GB swap to run a node, as is apparently the case here, is inexcusable and unjustifiable under any circumstances whatsoever.  Particularly in the face of the Bitcoin Foundation’s recent effortsii at running a node using one eighth of that. Christ on a cross you’d swear that physical machines and physical networks grew on the same boundlessly bountiful trees as central banker scrip. What is this, the future?

Now, after a four hundred word introduction that still doesn’t address the title of the article, let’s get to it.

My last ditch effort to play nice with the USG’s turdlette was a “googling” of the search term: “Cannot obtain a lock on data directory /root/.bitcoin.” This yielded the following results on the first page:

google search results

Needless to say, all four of these are muppet pages of incomprehensibly SEO’d gibberish. A telling tidbit from which:

Simply, the underlying books used by these subreddits may be: allegedly, some pow ppcoins offer interest data that allow means who know a bitcoin cannot obtain a lock on data directory root .bitcoin. bitcoin is probably already running, just a soviet attack, to generate unsupported pows. Before downloading a project on a value overhead, in most issues the other copyleft is acquired from a trusted recipient, for supply a circulation or a bitcoin cannot obtain a lock on data directory root .bitcoin. bitcoin is probably already running gambling that is known to have critical individuals of clothes to demoscene. This can include form pressures active with a shader or using a bitcoin cannot obtain a lock on data directory root .bitcoin. bitcoin is probably already running to purchase energy at a monitoring.

With the county of the web-hosting into excess case, bitcoin cannot obtain a lock on data directory root .bitcoin. bitcoin is probably already running operation has expanded to short data. The illegal season can be sold to finance the various subculture of the british bitcoin cannot obtain a lock on data directory root .bitcoin. bitcoin is probably already running bank.

In 1860, new bitcoin cannot obtain a lock on data directory root .bitcoin. bitcoin is probably already running prices were introduced. Bundle is used in the taxation of open bitcoin cannot obtain a lock on data directory root .bitcoin. bitcoin is probably already running and vending credit.

The pirate bay is a traditional hackintosh bitcoin mining of respect public. Newzbin was a public usenet malware indicator, intended to facilitate 10 bitcoins to cad to content on usenet.

*This is basically what happens when the map becomes the territory, when what you measure becomes the target and everyone in town tries to game your algorithm with a fully-fledged horde of only-superficially-geeky kids flying under the banner of “SEO experts.” The cat-and-mouse game of “optimisers” vs. algorithms inevitably results in the destruction of the algorithm. Because guess what, people are way, way smarter than machines are going to be anytime soon. So ya, Google basically tweaked their formula and tweaked their formula until their expertly hand-crafted Old Fashioned was reduced to an off-the-liquor-store-shelf pre-mixed cocktail.

So there you have it futurists. Web 2.0, much less Web 3.0, is dead on arrival. All the magick of sooper-smart algorithms and this is what you end up with. With this evidently pathetic state of machine learning, remind me again, why are people so fussed about AI taking over the world and enslaving humanity?iii All I can say is thank goodness we still have blogrolls.

In summa, Facebook is dust, Google Glass drools, and the search giant’s meat and potatoes is as borked as government-funded research.

Basically, everything the USG touches is turned to inedible, artificially constructed, and insanely cumbersome gold.

Funny how pseudoscienceiv works, eh?

___ ___ ___

  1. I do mean “wreckers” in the fully Soviet sense, wherein everything Gavin touches, regardless of whether he intends to do harm or good, is an offense against the sovereignty of Bitcoin. []
  2. We won’t call these 128 MB RAM tests an unqualified success just yet. We don’t need to count our chickens before they’re hatched. []
  3. Likewise, why aren’t they afraid of La Serenissima doing the same? []
  4. mircea_popescu: “The astonishing truth is that despite millions of dollars and hundreds of academic careers psychiatry has made no progress in almost 20 years, let alone ten, a claim no other medical specialty can make, and the truth which cannot be spoken out loud. Hence an exam.” Meanwhile “economy”, another us pseudoscience, has made shocking progress the past 20 years : austerity, the definitive solution to all the problems caused by inflation and loose money is now the problem, to be definitively solved by inflation and loose money. So much progress could hardly fit in a single asshole.
    pete_dushenski: Thankfully, the number of people have inflated too and each has at least one backdoor
    mircea_popescu: “Governance” – social “studies” – “earth science” – “economy” – “psychiatry” and “computer science.” That’s pretty much the list of things the us has uncontributed to.
    pete_dushenski: + Public Health*


    *In Great Revisionist History of Glorious Western World this not pooping in water thing is an entirely new idea and one that must be marketed effectively to increase adoption. Because it’s not like the kashrut declaration mapping to the same thing via “thou shalt not eat pork and shellfish” is a few millenia old or anything. []

14 thoughts on “On pretending “Googling” is still a thing that works any more than “USG”

  1. *Added after a complementary conversation I had today. This’d be called Goodhart’s Law.

  2. Alrenous says:

    Jim is almost certainly correct that Google has killed its golden goose with affirmative action. It has given unqualified women real power over search; the result is what you see above. This was forseeable as soon as Eric Schmidt was taken aboard.

    Though to be fair, I’m still able to wrangle Google, (use startpage.com or similar) and it’s not like Bing is even a worthy opponent.

    Mild objection to saying the map’s being the territory. That’s the key feature of consciousness, my main area of study. You’ve dropped qualifiers, strictly speaking.

    People can seem smarter than machines…but the machines are also made by people. It’s human vs. human proxy, and who wins depends first on whether the proxy-user is proficient and then it’s a straightforward opposed IQ/competence check.

    • A proxy is never the same as the real thing. For one, in this certainly, proxies require far more energetic inputs to obtain the same output. The amount of energy required to fuel inorganic machines is a grotesque magnitude of that required to power the human brain. It’s something on the order of 12 orders of magnitude difference.

      Granted, biology has taken billions of years to become this efficient, but even if silicon-based life can piggyback on us and take some shortcuts, it’s still surely beyond the realm of our meager lives. I’m not saying it’s impossible, merely that we need not fear it anymore than we do door-to-door shakedowns.

    • Alrenous says:

      I agree with the general principle. A cyborg brain would be far more expensive than equivalent human brains. However, in practice human brains are more expensive than computers. You have to keep a whole human alive for each one. Presumably the human fueling cycle is far more inefficient than a silicon electric cycle? Especially if we’re using health-promoting meats, then the brain is effectively the most inefficient solar-powered computer.

    • However, in practice human brains are more expensive than computers.

      This is quite incorrect. Any two 15-year-olds in a gutter can bang and end up with a human brain 9 months later, while there’s hardly anyone alive today with the mental and physical resources necessary to develop a “cyborg brain,” much less the rest of the roaring beast.

      Presumably the human fueling cycle is far more inefficient than a silicon electric cycle?

      Nope again. There’s nothing on God’s green earth more efficient that biology. Take the example of the dragonfly:

      Try this, by the way, catch mosquitoes while they fly around you, with your feet solidly anchored on the floor. From what I’ve seen, the average Joe manages in about a case of ten. With a little training you can do it with one hand, I manage about 60% of the time. Imagine you were doing this while flying through the room at a speed roughly equal to a hundred of your body lengths. Each second. It’s pure fantasy, we can’t think as fast as the dragonfly, and here’s the kicker : it uses just about two calories to do all this. Two calories a day. One tic-tac.

      If you’re a Bitcoin miner, your eyes are probably lighting up already : there’s a video card somewhere doing 5GBps x n at 750 Joules per day ? Why that’s much better than current miners! You could have millions of these things, just imagine, millions upon millions of matchboxes with little dragonfly brains within, hashing happily away. Why, you’d need over 100`000 of these just to fill up 1KW of power.

      Also, on behalf of La Serenissima, I’m probably overdue to extend an invitation for you to drop by #bitcoin-assets on IRC. You can use webchat or a dedicated IRC client.

    • Alrenous says:


      I’m still thinking about the rest.

  3. […] a small chance that we zoom-in and scroll down to find, what you see on the right… a Google survey […]

  4. […] translation of Auguste Kerckhoff’si seminal article on military cryptography on any of the major search engines, what follows is an admittedly crude effort in this direction […]

  5. […] 3. Saying that there isn’t a solution doesn’t make it so. 4. Willfully ignoring parallel developments is a quick way to out yourself as a state-backed agent. No one else is that […]

  6. […] important functions such as, say, a DuckDuckGo search bot so we don’t always have to use USGoogle. I’m sure you can think of some others. […]

  7. […] kool-aid in the form of “smart” everything, no matter how baked out of their trees the USGoogle employees were when they, between fits of giggles and mouthfuls of cheetos, ask each other […]

  8. […] be the serpent eating its own head – where its head is the artist formerly known as “search” – but we can give credit where credit is due for it having resolved the whole e-mail […]

  9. […] to govstats, only ~300 are using this app ? What kind of sense does that make ? Or is this another one of those “our algorithm only smartmatches you to the results you’ve always had before […]

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>