markonen

Crash is the best-case outcome for a browser bit flip. Here's bitsquatting:

https://ripe92.ripe.net/programme/meeting-plan/sessions/112/...

If you see photos/videos of Hetzner datacenters, their servers are essentially plain, low-end motherboards (1G network, few slots) sitting on shelves (don't mean that as derogatory, it's an efficient design). What it does mean is that their per-server costs are absolutely dominated by the very components that are exploding in cost right now: RAM, SSDs and (to a lesser extent) CPUs.

LC connectors are smaller and what the actual SFP+ modules typically have. If you want to run a link with just one fiber, you need BiDi optics.

FS does custom multi-fiber cable assemblies too (beyond the duplex patches which is basically the standard), and they can also include pull eyes on them if that’d be helpful.

Single mode is a good choice, common wisdom used to be multimode for short runs but the single mode stuff is not much more expensive and the standard 10km optics will likely brute force the signal over any mistakes like cable kinks or dirt on the connectors.

There’s nothing hard about it if you can run pre-terminated patches. Which you typically can since the connectors are so small.

What a blast from the past. I added that!

TIL! My problem with them requiring sqlite was that I assumed it would make a high availability setup either hard or impossible. Maybe that's not true, but definitely off the beaten path for headscale.

Apparently they've deprecated Postgres support and now only recommend sqlite as the storage backend. I have nothing against sqlite but to me this looks like Tailscale actively signaling what they think the expected use of headscale is.

Just had my first uncorrectable memory read error on our servers in 10 years or so today (in Sacramento). I'd like to think it's related because the alternative (buying new DIMMs) is too horrifying to contemplate

In the performance tests they said they used "consensus among 64 samples" and "re-ranking 1000 samples with a learned scoring function" for the best results.

If they did something similar for these human evaluations, rather than just use the single sample, you could see how that would be horrible for personal writing.

From the PostgreSQL 17 Beta 1 announcement:

> PostgreSQL 17 adds a new connection parameter, sslnegotiation, which allows PostgreSQL to perform direct TLS handshakes when using ALPN, eliminating a network roundtrip. PostgreSQL is registered as postgresql in the ALPN directory.

I'm looking forward to being able to offload PostgreSQL TLS to a standard (non-pg-specific) proxy.

They’re probably amplifiers rather than repeaters. Optical amplifiers don’t need to decode the signal to work. Here’s Wikipedia on erbium-doped fiber amplifiers:

> A relatively high-powered beam of light is mixed with the input signal using a wavelength selective coupler (WSC). The input signal and the excitation light must be at significantly different wavelengths. The mixed light is guided into a section of fiber with erbium ions included in the core. This high-powered light beam excites the erbium ions to their higher-energy state. When the photons belonging to the signal at a different wavelength from the pump light meet the excited erbium ions, the erbium ions give up some of their energy to the signal and return to their lower-energy state.

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

I moved away from Cloudflare—to self hosting our network infrastructure—because, while this didn’t happen to us, I was very aware that it could. We had a great deal on Enterprise for a couple of years, but zero guarantees that it would last (and some indications that it wouldn’t). I wanted to stop praying that they wouldn’t alter the deal.

Do you mean to imply that cloud services at higher levels of abstraction are cheaper per unit of compute than simple VMs? I believe you’ll find that the opposite is true.

At the scale discussed here, there are no free lunches.

Not really, because Concorde died in the seventies when the lie flat business seat didn't exist.

BA and AF managed to keep the zombie fleet going very profitably all the way until the end in the early 2000s, and that business wasn't killed by the lie flat business class seat either. It was killed by the impossibility of continuing to operate a tiny fleet of '60s planes forever.

Now if you said that the reason we don't have ANY supersonic passenger jets today is because lie flat business seats are good enough, then that's a more defendable position, but I'd still say that the overland flight restrictions limiting any SST to just a couple of routes is a bigger factor.

When I flew on Concorde the one thought I never had was "I wish I had a lie flat seat and half the airspeed".

Not sure why you're getting downvoted, this was definitely a key factor that made Concorde into a niche product.

It's not that customers preferred slower and cheaper flights over Concorde—they didn't, Concorde had very healthy average occupancy rates and operating the flights was very profitable for BA and Air France (they got the planes for free, of course).

It's that you can't fly a 1960s plane forever and you also can't amortize the design and development cost of new models with the only addressable market being first class customers travelling between the East Coast and a couple of European capitals (and this was directly caused by the overland flight restrictions).

Flying Concorde is one of my fondest memories :/

If you need macOS on the server for whatever reason, your only option is the Mac.

I have four Mac minis racked up for this, with two use cases: 1) iOS CI/CD and 2) some computer vision stuff using Apple’s Vision framework.

No complaints, but I obviously wouldn’t run anything that didn’t need to be on a Mac on these systems.

BTW, you can rack mount twenty Mac minis in the footprint of a single rack mounted Mac Pro (there’s a 1U mount that takes two and they’re small enough to mount on both the front and the back of the rack). So 20 M2 Pros per 5U.

They’re of course not a sysadmin’s dream but they do tend to stay up and Ansible works fine.

Apple's policies for external purchases are hilarous. The only goal is to be punitive.

For the External Link Account Entitlement that "reader” apps can use to link to purchase flows off-app, Apple forbids offering IAP in the same app. Why? Because they think this will discourage adoption.

For the new StoreKit External Purchase Link Entitlement that other apps can use for the same exact thing, Apple requires an IAP alternative. Why? Because they think this, too, will discourage adoption.

I think the crucial flaw here for the IPMI/BMC access use case is the fact that the card requires the server to be powered on to function. So if you accidentally turn the server off it’s game over for remote access.

Torment Nexus?

The idea that even 10-20% of users would power their Pis with PoE seems wild to me. Seems like a much smaller niche to me (even though I'm one of those people).

Anyway, I have been wondering whether a headless-oriented SKU wouldn't make sense. Pay for the PoE BOM by jettisoning the video output.

I'm absolutely 100% certain that AWS (for example) wouldn't do that for you with the instance types that feature direct attached storage.

The Moon being a great example.

Have they, though, if marked to market? After the deposit outflows this week?

Both of the upstream routers are mine, there’s no difference to the routes they have available. (They’re connected both to Tier 1 transit as well as the major local IXP)

I’m just getting the default route from both upstream routers. They’re not diverse in terms of the internet routes they have, this is just for HA.

I rent a single ~600 meter fiber that connects my home to the network infrastructure of the company I run. The rent is $75/mo or so. I have passive CWDM muxes on the line to run two 10G connections to two separate edge routers.

My home router is a MikroTik CRS309—it’s about $250, fanless, has 8 SFP+ ports and advertises my home network blocks over BGP to the two routers (for HA).

The setup works great. The best part is how it’s small and fanless and fits inside the very small wall box the fiber terminates in.

These were HTTP/2 requests, not connections, so no per-request TCP negotiation. I'd say it's very easy to generate a huge number of requests like this, even if it's over TLS.

I've been trying to shoehorn mTLS into existing infrastructure of Postgres clients and servers. This would be trivial if Postgres used straight-up TLS; as it is, though, you need sidecars that understand the Postgres wire protocol and that limits your choices to things like pgbouncer (which will add not-insignificant overhead) or Envoy (where the Postgres wire protocol support seems alpha quality or so).

I vividly remember how Ari Vatanen won the 1987 event in a Peugeot 205 (I was in elementary school). On the heels of the rally success he went on to become the PR face of Big Milk for the next decade or so.

You can absolutely regulate towards the goal of affordable, instantaneous money transfers within the context of the traditional finance system. The EU has done just that with SEPA (Single Euro Payments Area).

Here's the SEPA instant credit transfer service (maximum duration of ten seconds): https://www.europeanpaymentscouncil.eu/what-we-do/sepa-insta...