Last year, I thought we still had time to design PQ auth systems. Now, based on the pace of progress and on statements like Google's, I believe 1. we need to finish rolling out PQ kex yesterday 2. we need to start rolling out PQ auth now 3. it's too late to ship any new non-PQ design or system

Hey #AtmosphereConf, I am flying in tomorrow evening! 🥳 I was going to rest after a week in Taipei, and then @knowtheory.net and @ngerakines.me got me excited about bringing tlogs (as seen in CT, Go's sumdb, Sigsum, Sigstore) to atproto... and now I have a lightning talk on Sunday 😅

I am extremely excited by all the atproto apps, but I am a bit uncomfortable with how many store data in the repo but don't surface in the UX the fact that you can see other users' data. You and I know that all data is public (for now!), but the average user needs the UX hint or very explicit text.

Here's something weird: this post is the reply.parent of 4 posts, but also the reply.root of 2 of those posts, despite having a different reply.root itself. pdsls.dev/at://did:plc... The result is that bsky.app shows 4 as the reply counter, but only shows 2 replies. Any idea how this happened?

So uh, apparently $288/year is not enough to run a Mastodon service for ~3 users, and this is a known issue (masto.host/mastodon-con...) with no solution (github.com/mastodon/mas...!

omg a tiny MagSafe eReader, instant buy I can think of so many weird uses. Speaker notes for presentations. Read while on EMT shift. Motorcycle route reel. Quick reference notes for offline or no-camera activities. Fallback electronic pilot bag. If WiFi sync allows it, afk task status.

They sell entropy sources at FamilyMart in Taiwan in convenient 51-bit packages!

A great example of how the Google #golang team put thousands of lines of ultra-technical code in service of an incredibly easy-to-use and useful tool. If you ever deprecate code, you really should know about the go:inline directive!

Growing convinced we could and should ship new version cooldown in the Go modules ecosystem. The subdb and MVP get us far, and supply chain attacks are not much of a thing in Go beyond typosquatting, but we want to stay ahead of them. github.com/golang/go/is...

Find out how the source-level inliner in Go 1.26 can help you with API migrations. go.dev/blog/inliner

Dustin Moody from NIST: “you don’t need more than 128 bits of symmetric keys for post-quantum security” #rwc2026 Say it louder, for the people in the back!

I'm really glad that 404 Media correctly called out the "Proton did not provide any information to the FBI" line as disingenuous. It's not a good sign when a provider is more interested in playing semantic games than realistically explaining their threat model.

Who thinks "clean room" is needed to reimplement and put it into a new license does NOT understand copyright. Clean room is a trick to make litigation simpler, it is not mandated by law: rewrites are allowed. The new code just must not copy protected expressions. Linus was Unix-aware.

c'mon Safari

did something very silly, may have some at gophercon this year if you ever sent us a vulnerability report or contributed to Go crypto (or are just nice to me) thanks to @ljamesart.bsky.social who did the great art!

TIL about the git fast-import textual format! Lets me write tests for the new c2sp.org redirector against a synthetic git repository I can easily edit, and even gives me stable shorthands to refer to commits.

Passkeys will make something true that my father already asserted for 20 years: "I never had a password for this website, it just worked". 😄

The more nerds (I identify as a nerd!) complain about passkeys, the more I’m confident in their success. Just read a comment objecting to a passkey being automatically created for them. “Dark pattern!!1!1 Should be optional!!” No regular user wants to think about auth.

ssh is an obscure but widely-deployed command. It stands for Secure Snake Home and was made in the 90s to securely play snake online I made a massively multiplayer backend for it with support for thousands of concurrent snake players ssh snakes.run to join!

@aumasson.jp this might be a stupid q but, why does BLAKE tweak the ChaCha perm. into a block cipher to use it in M-D (where the message block is the cipher “key”), instead of using the perm. in a sponge?

You should be able to set up a fuzzing loop for HTML sanitizing libraries where you do setHTML on s and sanitize(s) and then compare the DOM, to find places where your sanitizer disagrees with the actual browser parser. I wonder if it would have tolerable false positives.

The Sanitizer API landed in Firefox 148, along with element.setHTML(). This lets you fully configure how HTML strings are cleaned as they're parsed. hacks.mozilla.org/2026/02/good...

the full PLC story relies on: - independent governance - public, formal auditability to stake its reputation - in case of emergency, credible exit even in the case described below, there is a story for auditability (see thread). plc read replicas shipped, governance and transparency logs ahead!

Dependabot security alerts have terrible signal-to-noise ratio, especially for Go vulns. That hurts security! Just turn it off and set up a pair of scheduled GitHub Actions, one running govulncheck and the other running CI with the latest version of your deps. Less work, less risk, better results!

We've released a self-hostable reference implementation of a PLC read-replica service to make the Atmosphere ever more decentralized. @retr0.id has details on what this all means. atproto.com/blog/plc-rep...

I just published a small security fix for a rarely used API in filippo.io/edwards25519. v1.1.1 is v1.1.0 + fix, while v1.2.0 has a few other changes. If you get notified but don't use MultiScalarMult, consider switching to a vulnerability scanner that respects your attention, like govulncheck!

Very few people know that the Go standard library has an internal DSL used to make sure that no new intra-stdlib dependencies are added when a change happens: github.com/golang/go/bl...

I wish that those surveys so often cited by InfoSec pundits that ask Do you fully trust AI output? Do you always verify AI output? also asked Do you fully trust your colleagues' output? Do you always verify your colleagues' output? Just to have comparative numbers, you know.

Alright, here we go, crypto/mldsa API proposal with external μ, deterministic signatures, signature context, and pure seed encoding. github.com/golang/go/is...

That theory is inconsistent with the history that Apple has (almost) always stopped providing security updates (or install signing) for older iOS N, very soon after iOS N+1 was released, except on devices unable to run iOS N+1. IIRC iOS 17 -> iOS 18 is the only exception in the last ~10 years.

Ah fuck, I didn't realize iOS 18 was not getting security updates on recent devices. I might have to put up with all that gratuitous transparency. At least macOS 15 still does if you click the right boxes.

A bare metal Go unikernel reference for AMD SEV-SNP execution. Starts under UEFI boot services with UEFI networking or can terminate them and connect through VirtIO networking (or soon gVNIC). github.com/usbarmory/tamago-sev-example

So, I hope it goes without saying for almost everyone, but don't do shit like this. The first issue mentioning a fork was totally fair, spamming every thread in the issue tracker without permission from the maintainer is not.

The Go Checksum Database guarantees that every single Go build uses the same source for a given module version. However, GitHub might not show you that code. pkg.geomys.dev displays the canonical source of a Go module, and it comes with browser extensions to replace pkg.go.dev source links.

Go 1.26 is out, and the announcement says: "Over the next few weeks, follow-up blog posts will cover some of the topics in more detail. Check back later." So you can wait a few weeks OR you can read my interactive Go 1.26 tour right away: antonz.org/go-1-26

I worked on implementing the runtimesecret experiment in this release. I'm very excited to see what snags people find using it, because I believe it is the first zeroization scheme to deal with unintentional stack memory spill and somehow, also the easiest one to actually use.

“Highlighting syntax is not always the most important thing to us. The information we want from code depends on what we're trying to do.” buttondown.com/hillelwayne/...

After seeing so many comments on here about it, I played through Firewatch in one sitting yesterday night. Went in blind, I don’t regret one minute of missed sleep. (It’s $4 without DRM. www.gog.com/en/game/fire...)

I've always wanted better tools to read and navigate—not edit!—source code. For example, it's incredible to me that there is no iPad app that does LSP features plus bookmarks, navigation tree, etc. Maybe now that LLMs made it more people's job to read code, we'll get good code reading tools?

Apps that get DNS verification functionality wrong really cause a lot of pain for their users. Having explained the problem a number of times, I finally decided to write it down so now I can just share a link.

in-band signaling is always a mistake and unix tooling is full of in-band signaling https://mas.to/@zekjur/116022397626943871

Just had the next Geomys CT log hw delivered, this one to be racked in an EU DC. This means my home office now has an Ampere Altra 64-core NAS with 96 TB HDD, a Dell PowerEdge R6515, a Milk-V Jupiter RISC-V 64-bit, redundant fiber, a Turris Omnia, a USB Armory, and an Enigma replica. And a MacBook.

I just bought a whole new server for the Geomys transparency services (CT log, tlog witness with SLA, more soon) using “PayPal Check out” and for some reason it was extremely funny to me.

I replaced eight different 5V, 12V, and 20V power supplies (for network gear, ext. hard drive, Milk-V board, air monitor) with two large USB-C ones (and various USB-C PD trigger cables). It might be the most satisfying thing I've done all year.

Everybody asking for ML-DSA in Go:

Installed the Fedora Remix from www.fedoravforce.org on my new Milk-V Jupiter (yes, I have a thing for weird computers), saw a message about skipped PGP checks, and... baseurl=http://openkoji.iscas.ac[.]cn/pub/... gpgcheck=0 No HTTPS either 💀😬 any network attacker can install whatever they want.

Much ink has been spilled on spam security reports, but in 2026 I expect security teams will be overloaded by VALID security reports, as AI tools clear the backlog of findable issues. You can already see an uptick in Go cryptography security fixes. aisle.com/blog/aisle-d...

ATProto devs did you know you can use selfhosted.social for users to create brand new accounts on the atmosphere in your applications? Both deckbelcher.com and blento.app uses our PDS. A user's journey into the atmosphere does not always have to be started from Bluesky. It can start from your app.

Because not using AI tools for what they excel at produces less secure code. For example, they are great at debugging (words.filippo.io/claude-debug...), they can find real issues in code review, they know more math than me, and they can write static analyzers I would have never had the time for.