Hey, it's HighScalability time:

 

• 900 GB: data stolen in Cellebrite hack; 99.24%: users identified by cross-browser fingerprinting; 72%: intend to migrate to a hybrid cloud; 90%: Google & Facebook ad traffic is useless; 5.2 terabytes per second: data from Australian Square Kilometre Array Pathfinder; 10 billion: searches on DuckDuckGo in 2016; $330m: Amazon's loss on Alexa; 


• Quotable Quotes:
  • @brucel: Breaking: Programmer accused of writing unreadable code refuses to comment.
  • @asymco: Remember Android first? App Annie believes the Apple’s App Store produced about twice as much revenue as Google Play
  • @bridgetkromhout: Describing your old-timer ranting as "greybeard" just makes me want to fight you with sed & awk at twenty paces. Be there tomorrow at dawn.
  • @StevenShorrock: Root Cause Analysis is: * Acceptable for simple systems * Inappropriate for complicated systems * Ludicrous for complex systems
  • @swardley: Five years ago Amazon was worth about half of Walmart, today Walmart is worth about half of Amazon.
  • Eric Raymond: In practice, I found Rust painful to the point of unusability. The learning curve was far worse than I expected; it took me those four days of struggling with inadequate documentation to write 67 lines of wrapper code for the server.
  • @swardley: past history shows many major players won't announce they're getting into the battle until some time after war has ended
  • @benthompson: Apple wasn't billed as phone maker / Amazon wasn't billed as infrastructure provider / FB wasn't billed as portal / Snapchat wasn't billed as TV
  • Jessitron: the biggest consideration in choosing whether to use libraries or services for distribution of effort / modularization is that choice of who decides when it deploys. Who controls which code is in production at a given time.
  • Hi Ben: The disruption of TV will follow a similar path: a different category will provide better live sports, better story-telling, or better escapism. Said category will steal attention, and when TV no longer commands enough attention of enough people, the entire edifice will collapse. Suddenly.
  • @leonidasfromxiv: I also don't understand why people compare Go with Rust. If you need a GC-less programming language: Rust; if you need a board game: Go.
  • Carlo Rovelli: The world isn’t just a mass of colliding atoms; it is also a web of correlations between sets of atoms, a network of reciprocal physical information between physical systems.
  • Chris Dixon: In the beginning, hardware-focused companies make gadgets with ever increasing laundry lists of features. Then a company with strong software expertise (often a new market entrant) comes along that replaces these feature-packed gadgets with full-fledged computers. 
  • Animats: The real question is "what do we do with a lot of CPUs without shared memory?" Such hardware has been built many times - Thinking Machines, Ncube, the PS2's Cell - and has not been too useful for general purpose computing.
  • @taavet: Very unfortunate that incumbents see tech only as a way to cut costs. Versus seeing tech to offer much better products.
  • NelsonMinar: This is what security looks like when your threat model is well funded government agencies.
  • Don Norman: The solution requires a different approach to the design of automation: collaboration. Instead of automating what can be automated, leaving the rest to the driver, we must develop collaborative systems so that the driver is continually involved in giving high-level guidance, thereby always staying active, always being in the loop. 
  • Thomas Frey: It took 50 years for the world to install the first million industrial robots. The next million will take only eight. Will this cause more jobs or few jobs in the future? I'm not convinced we know the answer.
  • @jtauber: "Every shot in Piper is composed of millions of grains of sand, each one of them around 5000 polygons."
  • rackforms: my point is the current situation, basically 2 companies controlling so much traffic, seems, well, bad for small business in this country. I value what they bring to the table and fully understand why they're so popular. But is things keep on this way where does that lead the guys like me? Is this just the way it has to be? Is the dream of the open Internet already dead?
  • @sheeshee: I think I know why it's called "DevOps" - "DevOops" was too obvious... ;)
  • greenspot: The open solution to a faster mobile web would have been so easy: Just penalize large and slow web pages without defining a dedicated mobile specification. That's it. This wasn't done in the past, slow pages outperformed fast ones on the SERPs because of some weird Google voodoo ranking, heck sometimes even desktop sites outperformed responsive ones on smartphones. If they had just tweaked these odd ranking rules in way that speed and size got more impact on the overall ranking there wouldn't have been any reason for AMP—the market would have regulated itself.
  • Juergen Schmidhuber: General purpose quantum computation won’t work (my prediction of 15 years ago is still standing). Related: The universe is deterministic, and the most efficient program that computes its entire history is short and fast, which means there is little room for true randomness, which is very expensive to compute. What looks random must be pseudorandom, like the decimal expansion of Pi, which is computable by a short program. Many physicists disagree, but Einstein was right: no dice. There is no physical evidence to the contrary


• RethinkDB is shutting down and here's the post-portem. Lessons: the database market is like Mad Max fighting in the Thunderdome; it's better to optimize for useless microbenchmarks than it is to be good; optimism isn't a strategy.


• Apple isn't alone in using custom hardware to thwart nation state level attackers. Google Infrastructure Security Design Overview. Good overview at Google reveals its servers all contain custom security silicon. Google designs "custom chips, including a hardware security chip that is currently being deployed on both servers and peripherals. These chips allow us to securely identify and authenticate legitimate Google devices at the hardware level."  Google encrypts data before it is written to disk, to make it harder for malicious disk firmware to access data. Google uses automated and manual code review techniques. Google uses automated software and code reviews to detect bugs in software its developers write. Google scans user-installed apps, downloads, browser extensions, and content browsed from the web for suitability on corp clients. Google uses a custom version of the KVMhypervisor. Good discussion on HackerNews, where a lot of the comments are on how Google needs this level sophistication to evade the prying eyes of governments.


• What happens when you embed machine learning into a DBMS in order to continuously optimise its runtime performance? You get Self-driving database management systems. Humans suck at tuning databases so this is just one more job AIs will eventually toss into the dust bin of history. TensorFlow was integrated inside Peleton training two RNNs on 52 million queries from one month of traffic for a popular site. Does it help?: early results are promising: (1) RNNs accurately predict the expected arrival rate of queries. (2) hardware-accelerated training has a minor impact on the DBMS’s CPU and memory resources, and (3) the system deploys actions without slowing down the application. 

Don't miss all that the Internet has to say on Scalability, click below and become eventually consistent with all scalability knowledge (which means this post has many more items to read so please keep on reading)...