lima

Google's SwitchTo kernel patches are similar to what you describe: https://lore.kernel.org/lkml/20200722234538.166697-1-posk@po...

Slides: https://web.archive.org/web/20200802205544/https://pdxplumbe...

Could just steal the key at that point.

Only if the key is generated off-key

The hard part is not backing yourself into a corner with the feature tree over time. Takes a lot of practice to avoid, and that's something AI could help with. Meanwhile, what their demo does is trivial.

Agreed, sync modelling in Solid Edge and NX can be really powerful, especially when requirements change unexpectedly.

This is data that criminals already publicly released.

Intent matters.

Thank you for this, I learned a lot from this comment.

The RADOS K/V store is pretty close. Ceph is built on top of it but you can also use it as a standalone database.

Lots and lots of heavy machinery uses Windows computers even for local control panels.

The problem is the long tail. Neural networks are not great at handling edge cases and it's not clear whether they will ever be.

What about consumer apps in Local Data Mode?

GitHub org permissions aren't very granular, especially with an org that is tied to enterprise SSO. Makes it hard to grant certain permission to outside members.

OpenVZ, Virtuozzo and friends definitely weren't secure the way gVisor or Firecracker are. You can still do that and some providers do, doesn't make it a good idea.

The funny thing with k8s is that Google doesn't use it (except GKE, and there's a reason it's one cluster per customer).

Their internal tooling scales just fine, but all it shares with k8s is some of the underlying concepts. Unlike, say, Bazel, gVisor or Gerrit, which are the real thing (minus some secret sauce tied to internal infra). k8s is good software, and best-in-class when it comes to open source options, but the idea that it is "open source Borg" is silly.

As a customer of Equinix Cloud... No thank you. Infrastructure vendors are terrible software engineers.

And in hammocks! They don't touch the ground and therefore don't count.

Much of kitty is written in C, particularly the (very fast) rendering pipeline.

The risk with such rewrites is ending up with a Python 3 situation and an ecosystem split. Sounds like YARA-X is (mostly) a stricter subset of YARA, and it's easy to write rules that are valid for both:

https://virustotal.github.io/yara-x/docs/writing_rules/diffe...

Although I wonder how long it'll stay that way? It'll be very tempting to add new features to YARA-X that won't be backported to YARA.

Yara is pretty much the industry standard for detection rules now.

Highly targeted attack on a specific machine. Probably not a threat actor but a red team or some other security researcher.

Use Clickhouse - it's specifically built for this use case (low latency OLAP).

It archives a much smaller part of the internet, it't not an IA replacement.

Why would the U2F-on-Ledger route stop working?

By calling them

Talos appears to rely on existing components to a somewhat greater degree in comparison: https://github.com/siderolabs/pkgs

There aren't even any poor security practices. It's just security theatre and the DOC is overreacting.

Not in terms of specific threat model.

Phabricator is what we used before we migrated to Gerrit. The review UI was very nice, the issue tracker is still best-in-class, but Arcanist was a menace. Much prefer a simple "git push" over that client-side PHP contraption :-)

Are there plans for Gerrit interop with Jujutsu?

There isn't, you need to contact them so they can delete the offending objects.