cfstras

In this case, IIUC the swap-restore mechanism just can’t use fragmented files because it runs before filesystem drivers can do their work. So it just gets an offset and probably derives the length from a header stored there, using raw disk accesses to thaw the RAM contents.

In general terms though, I hear disk fragmentation still makes a difference, f.ex. because disk page sizes don‘t match filesystem extent sizes, and disk optimizations like prefetching or even garbage collection work better with contiguous data.

That’s my point, it’s different for each distro & partitioning setup, so I wouldn’t know what to share here. Having a swap partition makes it as easy as adding `resume=/dev/sdXYZ` or `resume=UUID=…` to your kernel parameters.

swap files make setting up hibernation a bit more complicated - in most cases the file has to be contiguous, e.g. you have to defragment it once after allocating. Then, you need to tell the bootloader the byte offset of the file on the partition.

It‘s going to be a godsend for people with M1 macs when Apple or App vendors decide to drop support for them.

The hardware is super capable and will make a great fileserver, TV box or HomeAssistant instance once I‘m done using it as daily driver.

Similarly, because of distros like Fedora and Ubuntu, I can keep using a 15y old Samsung notebook with a modern firefox, whereas Windows 8 already broke Samsung‘s display brightness driver and thermal management, and 10&11 are entirely unusable.

Demand would go down, so meat companies would reduce breeding to reduce output. Or start an ad & lobbying campaign to increase demand again.

A couple years ago, I was at a station waiting for a (delayed) ICE train. I couldn't buy a ticket at the machine or with the app, since the train had already departed (if it had been on schedule). The ticket machine also wouldn't take VISA / MasterCard, only the more common Girocard (most people still call it EC)

Later, in the train, when I asked the conductor to buy a ticket with my Girocard, he said "That's not a commonly used payment method" and asked for VISA, or cash (not having any to provide change, obviously).

If my math is right, that gives you 83 clock cycles to calculate a single sample. on a 16 core, theoretically 1333 cycles. that‘s not a lot, considering you don‘t nearly 100% of the cpu all the time.

That same graph had me jump towards the sampling theorem - playing back an animation with linear interpolation creates hard edges, e.g. frequency spikes. I‘m not sure if the movement space is comparable to audio here, but I can‘t see why not.

so; if the sampling theorem applies; having 2x the maximum movement „frequency“ should be enough to perfectly recreate them, as long as you „filter out“ any higher frequencies when playing back the animation by using something like fft upscaling (re-sampling) instead of linear or bezier interpolation.

(having written this, I realize that‘s probably what everyone is doing.)

Most languages with macro or templating should be able to define this behavior as a library. Rust and C++ come to mind, for example.

is this satire? I honestly can’t tell.

Fellow German here, I’ve experienced a total of ~2 hours, in one single outage, about 20 years ago. Not even a single brownout since.

Hmm… That makes sense. Though if you are building from source, the only gain you have over regular go tooling is downloading from your favorite mirror rather than github&co. I hope it doesn’t install these source-packages for prebuilt binaries…

What are the golang- packages about in OpenSUSE? Go doesn‘t support dynamic linking (except for non-go-pieces like glibc or via cgo). Do they contain source-code? Or pre-built .a‘s? Both would have a hard time not conflicting with go‘s own build tooling, right?

If you want to use this as a toy, sure. But please don’t rely on it for messages you want to actually get delivered: These gateways frequently drop messages for random reasons, or block your IP if you send more than a few messages at a time. Also, in countries other than the US, expect the gateways to be long gone, or completely unreliable.

> The price for a replica starts at 0.001 ETH. Each time a particular NFT piece is replicated, its next replication cost will double. So be sure to replicate popular pieces early!

So who gets the replication fees? Is this just a cash grab by the authors of the site?

I cannot stress this enough:

    -webkit-font-smoothing: antialiased

Looks _horrible_ on non-retina (read: most external) monitors with macOS. Browsers have sane defaults for every OS! Please retain them rather than making me install TamperMonkey just to read your blog.

I'm not convinced congestion pricing will actually fix the problem. Once public transit exists and is more useful than your own car, people will use it.

I consistently notice this when visiting larger cities in Germany. Good public transit => I'll take the bus/tram/subway every day. Slow or unreliable => I'll gladly take an Uber or buy my own car.

Maybe surge pricing could be use to finance investments into public transit. Just make it a function of the horsepower-to-people ratio. Anything below 90HP/person is free. If you can afford a nice car with lots of power and don't carpool, pay up.

You could probably use something like https://github.com/benbjohnson/litestream to backup your signal db.

Yes, as they require a significantly less amount of stuff you need to deal with. WINE needs to implement vast amounts of Windows APIs to make something like .net work, while cosmopolitan limits you to low-level C (and things built on it, like lua) at compile time.

Java has this since 8:

    Arrays.asList("hello", "world").parallelStream()
        .forEach(s -> System.out.println(s));

If you want to calculate stuff, you can

    List<String> suffixed = Arrays.asList("hello", "world").parallelStream()
         .map(s -> s + "_suffix")
         .collect(Collectors.toList())

ah, that makes sense. Looks like the gdate I'm using only links against libsystem, and I don't have a glibc anywhere. So it's the darwin libc that doesn't have this bug.

Testing with https://hub.docker.com/_/busybox, it seems like musl also has this problem. uclibc is not affected.

The part about "using the text of the invalid timezone" seems to be fixed in latest GNU date (8.32):

  $ TZ=EDT gdate
  Wed Feb 10 13:22:23 UTC 2021
  $ gdate --version
  date (GNU coreutils) 8.32
  ...

8.30 (shipped in debian buster) still has the problem.

The BSD version of date shipped with MacOS 10.15 also does not seem to have this bug.

Docker on Mac and Windows both use virtual machines to run a Linux OS, which then runs Docker normally.

Since the 10,1 is not selling yet, where are these numbers coming from? Are Apple employees publishing benchmarks? Or are randos on the internet just uploading fake numbers?

- I was looking for "desktop-solutions" comparable to Zoom, so WhatsApp and Telegram are out of the question (Telegram doesn't do group calls AFAICS).

Some notes:

- Wire has published a detailed whitepaper on e2ee. https://wire-docs.wire.com/download/Wire+Security+Whitepaper...

- Jami (formerly GNU Ring) has an interesting post about having e2e here: https://security.stackexchange.com/a/162603/243716

- Jitsi e2e is testable, but they note that it's not completely finished. Key exchange has to be done manually.

- https://github.com/bigbluebutton/bigbluebutton/issues/9893 suggests BigBlueButton don't have e2e yet

- Wikipedia claims "no encryption protocol" for OpenMeetings https://en.wikipedia.org/wiki/Apache_OpenMeetings

---

So that leaves Wire and Jami. Thanks for the info!

What other popular group video meeting tools are e2e? I know of none.

I remember reading a while back that Zoom claimed a few times they were e2e-encrypted, but what they meant was transport encryption.

People have also found that you used to be able to discover the other routers in your „Carrier-Grade-LAN“: https://media.ccc.de/v/32c3-7133-beyond_your_cable_modem

CG-NAT, more specifically DS-Lite, is being used by lots of ISPs in Europe. In Germany, it has been the default for new consumer contracts for ~6 years now.

The mumble protocol is a nicely documented example for using encrypted Opus (OCB-AES128) over UDP: http://mumble-protocol.readthedocs.io/en/latest/voice_data.h...

If the qualifying process is long and expensive, doing a one-off fix does seem more logical. This is assuming you can be sure the fix really is isolated. I do feel like a lot of companies over-exaggerate the need for multiple supported releases.