Article note: It's a 2018 essay and _way_ too wordy, but the essential argument - that super-optimizing for present conditions will leave an entity without the energy/flexibility to adapt to changes in conditions - is good and, although everyone would say they agree if asked, there is a lot of personal and organizational min-maxing behavior that indicates they behave to the contrary.
Article note: Pressure Swing Adsorption is just neat.
I'd been curious how hard it is to DIY one of these things since Zeolite 13x is pretty cheap and accessible, the concept is simple, and the pressures aren't _alarmingly_ high. Looks like the OxyKit folks saw that thought through.
I'm kind of curious how - in the current post-COVID in the US market context - DIYing this thing competes with a cheap commercial one (which looks like it can get down into the $200 range) or maybe an electrolysis splitter on cost and practicality.
I'm curious about some applications for a little concentrated O2, like the potential of using an O2 assist on a cutting laser.
Article note: WTF is arm up to here? Qualcomm ate Nuvia whose architecture license apparently specified they were for server or non-mobile or something applications, and are now repurposing a derived core for the laptop market... so ARM is suing to cut off the competing core. It's possible Qualcomm ate Nuvia as a side-step specifically _because_ ARM was being weird about giving them an architectural license.
I assume they just want a bigger licensing cut, like everything they've done since Softbank.
Article note: Aw. Her track record is one that indicated a truly unusual level of genius;
Even before widespread knowledge of her work at IBM on what was probably the origin of Superscalar architectures (because it was pre-transition), her VLSI work was world-changing.
Article note: This is an _incredible_ story of google being structured to incentivize greenfield development instead of maintaining existing products.
I have a general distrust of phones so I've only occasionally used any of these products for P2P features, but the level of bungle is into "you couldn't make this shit up" corporate ineptitude territory.
Google has killed off the Google Pay app. 9to5Google reports Google's old payments app stopped working recently, following shutdown plans that were announced in February. Google is shutting down the Google Pay app in the US, while in-store NFC payments seem to still be branded "Google Pay." Remember, this is Google's dysfunctional payments division, so all that's happening is Google Payment app No. 3 (Google Pay) is being shut down in favor of Google Payment app No. 4 (Google Wallet). The shutdown caps off the implosion of Google's payments division after a lot of poor decisions and failed product launches.
Google's NFC payment journey started in 2011 with Google Wallet (apps No. 1 and No. 4 are both called Google Wallet). In 2011, Google was a technology trailblazer and basically popularized the idea of paying for something with your phone in many regions (with the notable exception of Japan). Google shipped the first non-Japanese phones with the feature, fought carriers trying to stop phone payments from happening, and begged stores to get new, compatible terminals. Google's entire project was blown away when Apple Pay launched in 2014, and Google's response was its second payment app, Android Pay, in 2015. This copied much of Apple's setup, like sending payment tokens instead of the actual credit card number. Google Pay was a rebrand of this setup and arrived in 2018.
The 2018 version of Google Pay was a continuation of the Android Pay codebase, which was a continuation of the Google Wallet codebase. Despite all the rebrands, Google's payment apps were an evolution, and none of the previous apps were really "shut down"—they were in-place upgrades. Everything changed in 2021 when a new version of Google Pay was launched, which is when Google's payment division started to go off the rails.
Article note: It's cute but... isn't the major threat model that makes these things a bad idea that an attacker could technically or legally obtains access to the logs, which is still there if its all local and open source?
Article note: 1. Not unexpected, but if Apple can veto stuff in 3rd party app stores, they aren't meaningfully independent, and this really shouldn't count as complying with the DMA.
2. An iPad with a keyboard case that could run VMs would be a rather compelling device, computers-with-touch and/or stylus tend to be a little half baked, but iOS is too restrictive to be a competitor on its own.
Article note: This is a way more interesting piece than I expected when I first saw it go by, it goes into the full history and explores where the common wrong answers came from, and the modern refinements to deal with computers being way more complicated and dynamic than they once were.
I've actually refined my mental explanation after reading, I was close but it's now: "PID 0 is the null process that gets scheduled when nothing else is ready to run. When running, it tries to not be running by ceding to another process or turning off resources."
The very short version: Unix PIDs do start at 0! PID 0 just isn’t shown to userspace through traditional APIs. PID 0 starts the kernel, then retires to a quiet life of helping a bit with process scheduling and power management. Also the entire web is mostly wrong about PID 0, because of one sentence on Wikipedia from 16 years ago.
There’s a slightly longer short version right at the end, or you can stick with me for the extremely long middle bit!
But surely you could just google what PID 0 is, right? Why am I even publishing this?
Article note: Mmmyep. When the person whose research was one of the main reasons we could manage a global pandemic says "I got marginalized out of academia because I wasn't all-in on playing prestige games" the answer is not "How dare you! Those prestige games are super important!"