praseodym

The vulnerability here is that regular non-administrator users can also read sensitive registry hives from the shadow copy. This allows for local privilege escalation exploits.

This makes me wonder whether 34PB (+1PB per month) of analytics data creates more value than it costs to store and process all of it. I’d think that storing only aggregated information would provide nearly as much business insight to make strategic decisions at a fraction of the cost.

There is an SSHFP DNS record for this purpose, which should be just as secure with DNSSEC enabled.

https://en.wikipedia.org/wiki/SSHFP_record

Saeco is a Philips brand, no funny rebranding business going on: https://en.wikipedia.org/wiki/Saeco

My local provider requires PPPoE over fiber but accepts any password. It’s probably some legacy artefact that’s too complicated to remove from the stack.

This still exposes whether you are online, just not the last seen time.

I think that’d definitely help. The Kubernetes project requires a maintainer to reply with /ok-to-test for the CI bot to begin testing a PR. This also helps reduces load on their CI systems, as the project gets a lot of commits.

Because I had to look this up: ‘SNS’ is an abbreviation for Social Networking Service that is mostly used in Japan.

The second example in the article shows someone using 5000 kWh in January of this year alone — how does one household reasonably use that amount of electricity? In The Netherlands and Germany the average is around 3500 kWh per household _per year_.

Probably malware using domain fronting techniques for C2 traffic played a role in that decision. E.g. https://threatpost.com/apt29-used-domain-fronting-tor-to-exe...

The annoying thing is that YouTube’s audio quality is linked to the selected video quality, so watching a music video in SD will often also result in inferior audio.

There are ESP32 variants with a dual-core Xtensa processor (https://en.wikipedia.org/wiki/ESP32).

The HomePod Mini also needs wall power, so it’s just as wireless as the Sonos speakers.

I guess with the lack of Dolby Atmos they would be more comparable to a Sonos Beam plus two rear speakers (e.g. Symfonisk) which can be bought for about $500. And that works with HDMI ARC for any input you might have on your TV, AirPlay, and every service that Sonos supports.

Post-it Page Markers or Flags work great as well and are already the correct size for most laptops.

It’s also called “komkommertijd” (“cucumber time”) in Dutch. Not pickled, because we call pickles “augurken”.

Cloudflare did write a blog about getting poorer performance with AVX-512 enabled in a mixed workload: https://blog.cloudflare.com/on-the-dangers-of-intels-frequen...

Apparently they are, Zerodium has even stopped accepting new iOS privilege escalation exploits: https://www.theregister.co.uk/2020/05/14/zerodium_ios_flaws/

SubjectAltNames should always include the common name, modern browsers won’t even look at the common name value any more (because it’s validation algorithm is not well defined, whereas it is for subjectAltNames).

IBM Z mainframes don’t use the POWER architecture, but have their own z/Architecture instruction set and CPUs: https://en.wikipedia.org/wiki/Z/Architecture

PIR’s response is at https://www.keypointsabout.org/blog/statements-in-response-t...

> Yes, this means you could theoretically make an analog keyboard which reports the level of key depression.

There is at least one company making analog keyboards, mainly for gaming: https://wooting.io/

Apparently a notification will show up when optimised charging is enabled. I’ve never seen it, because the functionality also requires Significant Locations and other location services (listed near the bottom of the page) to be enabled.

That seems to me to be quite unfriendly to privacy. And given that I always charge my phone overnight and use my phone’s alarm function, it could simply use that as the target time to finish charging to 100%. No complicated location-based ML needed.

Better yet, run `smartctl -t conveyance $DEVICE` first, which will run a short (few minutes) test to check for transport damage. `smartctl -a $DEVICE` also reports self test results, so always check those as well.

Siri/HomeKit was never supported on the Hue Bridge v1, only on the current v2 bridge.

I can’t speak for the Graviton2 CPUs, but AMD Epyc CPUs have RAM encryption with per-VM keys for increased isolation: https://developer.amd.com/sev/

Related, from the article:

> We hypothesized that increasing the Node maximum heap size from the default 1.7GB may help. To solve this problem, we started running Node with the max heap size set to 6GB [..], which was an arbitrary higher value that still fit within our EC2 instances.

Sounds like they were utilizing their EC2 instances very poorly. Why not run more workers per instance, or switch to an instance type with less RAM (or more CPUs)?

Debian arm64 works fine on the Raspberry Pi 3 (https://wiki.debian.org/RaspberryPi). The Pi 4 changed the boot sequence so that Debian doesn’t work any more, making this the first 64-bit OS for the Pi 4.

Microsoft has already heavily invested in getting Kubernetes running on Windows as a fist-class citizen: https://docs.microsoft.com/en-us/virtualization/windowsconta...

In the Moby issue you mention that you are using live restore (https://docs.docker.com/config/containers/live-restore/) which is most likely where the problem is. Docker daemon restarts, existing containers are kept alive, but the restarted Docker daemon doesn’t know about those existing containers yet and thus thinks their volumes are unused.