Tabs, Not Spaces

October 10th, 2020
Politics, privacy, and corporate maneuvering. Even in the tech world, you can tell November is just around the corner.

Download audio (Duration 07:45, 5.3 MB)

Show Transcript and Links
It's been quiet in the Linux hobbyist space, with just a handful of minor releases from a few different software projects, and new updates landing from some of the smaller distributions. But that's probably no bad thing, as there's been an awful lot of broader industry changes playing out this week — and now I've got the time to talk about them.

Starting with hardware, Arm has announced that it will be removing 32-bit support from its Cortex-A processor cores within the next two years. The processor range, which is typically used in high-end smartphones and laptops, is set to become 64-bit only when the company releases its CPU design code-named Makalu in 2022, ahead of its commercial availability the following year. The impact on developers of smartphone apps is likely to be minimal, as Google has long been phasing in stricter requirements for them to provide 64-bit versions of their code, and intends fully obsoleting 32-bit Play Store support by August next year — except for legacy devices that are incapable of running anything else.
Pushing the boundaries of performance and security to unleash the power of 64-bit computing
[Old] Android Developers Blog: Get your apps ready for the 64-bit requirement

Meanwhile, AMD has unveiled a new range of desktop CPUs based on its Zen 3 microarchitecture, which will be available to purchase from next month. The Ryzen 5000 series looks set to provide a speedup over its predecessor of around 20%, and should continue the recent trend that has seen AMD overhauling Intel in the CPU price-performance stakes. And like other major players in the silicon space at the moment, it looks as though AMD harbors ambitions to further extend its reach — and the Wall Street Journal is reporting that the company is close to acquiring FPGA manufacturer Xilinx. The deal is alleged to be worth around $30bn, and comes five years after Intel snapped up Altera, which at the time was the other leading light in the FPGA marketplace.
AMD Ryzen 5000 and Zen 3 on Nov 5th: +19% IPC, Claims Best Gaming CPU
AMD Is in Advanced Talks to Buy Xilinx
[Old] Intel Agrees to Buy Altera for $16.7 Billion

IBM has announced a corporate restructuring, which will see its Managed Infrastructure Services unit spun off into a separately traded public company. The new business will focus on designing and running infrastructure for its clients, which currently include more than three-quarters of all Fortune 100 companies, while IBM proper will double-down on the hybrid cloud and AI marketplace. But with IBM seemingly ready to slap the 'cloud' label onto most everything that it does these days, some questions have been asked about how profitable that side of its overall business really is — and there's more about that in today's shownotes. And as a fairly well-known angel investor recently put it on Twitter: "IBM splitting itself up into two companies is kind of like when Netflix proposed to split itself up into its future-oriented streaming business and its dying legacy DVD business. Except to make the analogy work for IBM, imagine Netflix had no streaming business..."
IBM Jettisons Legacy Services to Focus on Hybrid Cloud
IBM’s “Cloud” Business (or Lack Thereof)
Twitter snark from Charles Fitzgerald
[Bonus link] My first two predictions for 2020 — IBM and Trump

Google didn't exactly endear itself to other industry players including IBM when it decided to spin the Istio project out to a new organization clearly under Google's own control, earlier in the year. And while it's since partially changed the governance of that project, it looks as though the lesson was learned when it came time to unload its Knative project. Knative adds components for deploying serverless applications on top of Kubernetes, and it's being reported that Google is giving up its direct control over the project to a five-seat steering committee — which will have rules to prevent any single organization from holding more than two seats.
[Old] Why IBM doesn’t agree with Google’s Open Usage Commons
Google will give up direct control of the Knative open-source project

Google has also been in the news this week as its long-standing battle with Oracle over the Android Java APIs finally started to see some action in the US Supreme Court. And the consensus of opinion from most everyone who followed the proceedings is that the Justices seemed far more receptive to Oracle's oral arguments than Google's — which proved a bit embarrassing for a tech press that's largely spun the opposite story over the last few years. And perhaps to distract from their own failings in not seeing this coming, some journalists chose to hit out at Google's counsel — which is rather like blaming a doctor for failing to slow the spread of a malignant cancer, despite their genuinely best efforts. While the case hasn't been decided yet, the stakes remain high for the entire software industry, with potentially far-reaching consequences, no matter who eventually prevails. I'll bring you more as the story unfolds, but in the meantime check out today's links to see where things currently stand.
After ten years, the Google vs Oracle API copyright mega-battle finally hit the Supreme Court — and we listened in
Google’s Supreme Court faceoff with Oracle was a disaster for Google
Supreme Court inclined to affirm Federal Circuit's copyrightability holding in Oracle v. Google, possibly unanimously—fair use may be remanded
FACT CHECK: Ten falsehoods and fallacies Google's lawyer told the Supreme Court about Oracle's Android-Java copyright case

And United States law was also one of the main drivers that recently prompted a group of companies to come together and propose a new privacy mechanism for web browsers. When California passed its Consumer Privacy Act, it gave web users a slightly watered-down version of the protections available under the GDPR in Europe, and some control over how their personal information would be used by online companies. One of those rights was an opt-out of the sale of their data to other parties, but until now users have had to deal with individual web companies to exercise that right. Now, a group of organizations including Brave, DuckDuckGo and the EFF, along with backing from The New York Times and The Washington Post, are proposing a new HTTP header that users can easily set and forget, to make their intentions clear to website owners. And while the new Global Privacy Control isn't an industry standard just yet, a further push to see that happen might come as a result of the elections in November — as Californians will again be voting on a proposition to further increase their online protections, and close a loophole in the original privacy legislation.
California Consumer Privacy Act (CCPA)
Announcing Global Privacy Control: Making it Easy for Consumers to Exercise Their Privacy Rights
California Voters Asked to Amend Privacy Law

Having trumpeted the fact that its Privacy Badger tool would include support for the new Global Privacy Control, the poor old EFF had to announce a rather less welcome change shortly afterwards — after discovering that it's anti-tracking tool could actually be abused in order to track people using it. Having been tipped off by security researchers at Google, the EFF has now disabled Privacy Badger's local learning feature by default — meaning that unless users manually enable it, they'll simply receive a centrally-collated list of trackers to be blocked instead, which is a feature that plenty of browsers and extensions already provide.
Announcing Global Privacy Control in Privacy Badger
Privacy Badger Is Changing to Protect You Better

And let's wrap up today's politics- and privacy-heavy episode with a story that brings the two together. And it's yet another one that you might have missed, as the tech press has been strangely quiet about correcting the record on something else that they originally got desperately wrong. A few years ago, you couldn't move for stories about Cambridge Analytica, and how a group of dastardly data scientists managed to convince the mindless masses to vote for Donald Trump in the United States, and for Brexit in the United Kingdom. Now, a thorough three-year investigation by the Information Commissioner's Office in the UK has found not only that Cambridge Analytica didn't sway British voters — but that it couldn't have done so, as it didn't play any part in the Brexit campaign. The investigation included the examination of computers and servers seized by UK authorities, along with a review of over 300,000 company documents and 700 terabytes of data — and really can't be written off as the coverup that a handful of conspiratorially-minded journalists are now attempting to paint it as. The ICO review also found that while Cambridge Analytica did have numerous data points relating to the US population, it mainly acquired this data commercially, and wasn't doing anything significantly different from its competitors. But what was different were the overblown claims being made by the management of the company about its impact and influence, which caused concerns even among its own staff. So while Cambridge Analytica might have claimed to have some uniquely powerful technology for influencing voter behavior, it turns out that the people most influenced were a number of gullible journalists, who swallowed the company's marketing spin, hook line and sinker. And while it would be really tempting to mock anyone who got taken in by the Cambridge Analytica story, a more positive take-away might be to realize that journalists are often no better informed than the rest of us — and to treat everything that you hear in this polarized and clickbait world with a healthy dose of skepticism.
ICO letter about Cambridge Analytica

Send your tips, release news, and feedback by email, or reach out on Twitter or Mastodon

Want to support the show? Then tell somebody about it! The best way to secure the podcast's future is by having a large and engaged audience — so help out by sharing it with others.

Copyright © 2020 Tabs, Not Spaces   About | Privacy policy