Shurimal

140 dB under water is not the same as 140 dB in air. For underwater noise reference level is 1 uPa, in air noise level reference is 20 uPa. 140 dB under water would translate to 114 dB in air. Still impressive, a trained opera singer or a typical home hifi system can achieve somewhere around 105…110 dB, but far from a gunshot.

EDIT: in another article it was mentioned that the actual sound level is 108 dB at 1 m which would translate to 82 dB at 1 m in air. 1 m distance is the standard distance to measure the SPL level of an object, eg a loudspeaker. Far less impressive and very, very far from an actual gunshot that is ~140 dB at 1 m distance. Science reporting at its “best”.

Privacy and anonymity are two very different things. Meta et al very much tie your real world ID to your online persona plus track, analyse, store and sell everything you do on their platforms (and possibly what other sites and apps you use).

Your posts on Facebook can be set to “private” (only Meta and you/your friends see them), but they’re never anonymous. Eg if you start saying unwanted stuff if you live under an oppressive government, you will have IRL trouble.

Your post on Fediverse can be public (everyone can see them), but anonymous (no-one can tie your IRL identity to your online persona). Eg you can make an account here and say things about your oppressive government and if you use basic anonymity tools (VPN-s, e-mail aliases etc) you cannot easily be tracked down.

You can also be private and anonymous, eg tipping off journalists about stuff going on under your oppressive government by using anonymous, E2E encrypted message service like Briar. That way no-one can eavesdrop on your comms and only you know your real identity.

Leslie Fish is still writing music. Plus her old works from 1980-s (eg It’s Sister Jenny’s Turn to Throw the Bomb and Firestorm) are as relevant now as they were 40 years ago.

O7

On my server:
OpenMediaVault (NAS OS based on Debian)
Syncthing
Home Assistant
Zigbee2MQTT
Docker
Portainer
Radicale
Navidrome

On my phone:
Syncthing
Tailscale
Feeder
DAVx⁵
OSS Document Scanner
RPNcalc
DSub
EDSY

On my PC:
Odyssey Material Helper
EDDiscovery
EDSY
ObservatoryCore
Paint.net and GIMP
OpenRGB
Tailscale

That’s pretty much like Wild Weasel SEAD works. Get locked on by a SAM radar, lock your HAARM to that radar’s signal and press fire🙂

As it should be #DynamicRangeIsLife

The first HL had surprisingly good sound with great dunamic range. HL2 felt wimpy, guns never had the same oomph to them.

No, it isn’t. OS and app volume controls are not implemented on driver level, but in each application individually, or you wouldn’t be able to change OS volume and in-app volume independent of each other. It’s simple math, multiplying audio sample values with a coefficient, best done in 32 bit floating point.

The question is not whether to do the math at driver level or in the userspace. The question is: if the user sets their volume slider to the middle, what value that coefficient should be? Most apps use simple linear correlation (middle point halves values which is 6 dB of attenuation or -6 dB(FS)) which is not how human hearing works. Log volume control would have the middle point at, eg, -40 dB(FS) and zero point at -80 dB(FS), giving psychoacoustically useful range in both halves of the bar. This is how analog volume controls on amplifiers work (not exactly so, but pretty close).

Driver level volume control can be done, but then you’d need to open your sound card control app and set it there. It would be an addition to OS and app volume controls. It would not be tied to OS or in-app volume controls or affected by standard multimedia keys on your keyboard. And if you decide to do OS volume control at driver level, in-app volume controls would still need to exist and be at the mercy of the app devs competency at implementing it.

Implementing proper logarithmic volume controls and defaulting them to -20 dB(FS) would be great. But the math involved is slightly more complicated* than the simplistic “multiply everything with a coefficient between 0 and 1” so devs won’t bother (if they even know about logarithmic volume controls at all).

*I did logarithmic volume slider in Jscript for foobar2000 using a Jscript GUI plugin and it was not too difficult, but not straightforward either. Getting the button states and scaling to work correctly was more difficult and I never solved some annoying bugs. That was the first and the only “programming” I’ve ever done in my life.

Volume sliders never sound linear to me

Ironically that is because (with very few exceptions) every application from OS-s to streaming service webapps to games to mediaplayers uses linear volume slider. Human hearing is logarithmic.

The way typical volume slider works is multiplying the audio sample values with a coefficient that is ≤1. Ie, if you set volume to 50% the input is multiplied by 0.5 and as a result the signal voltage level on the analog output to your headphone or loudspeaker drivers is halved. The kicker—halving the voltage is just 6 dB less volume. This is why if you have sensitive headphones (or big, powerful speakers) you find that you have to keep the volume slider in your OS at 10% or even lower to not blast your ears off. And why the upper half of volume sliders is completely useless.

I have an unconventional speaker setup that makes classical analog volume control completely impractical. Since said setup has the maximum sound pressure level output of around 110 dB at full scale digital input, I have to keep the OS volume slider at 30% and in-app volume sliders at around 20%, resulting the total multiplier of 0.06 (or about -26dB full scale) to have comfortable volume levels. Only exception is Elite: Dangerous; with sound set to full dynamic range I can keep the main volume slider at maximum and enjoy glorious dynamics. Youtube is also surprisingly reasonable, probably because they normalize to -14dB LUTS or something similar.

The Mist.

Not movies, but Rifters series, Blindsight and Echopraxia by Peter Watts; and Killing Star by Zebrowski and Pellegrino. These will never become movies or TV, they’re just too nihilistic and have some extremely heavy themes. Watts especially does not shy away from describing and closely analyzing the psyche of some truly horrible characters in Rifters series.

Yes, because vast majority of orgs both in private and public sectors suck at securing their systems. Either:
-The admins lack the knowledge and skills to properly configure their stuff.
-The admins are not given the resources they need to update and secure the systems.
-The in-house parts of the system rely on some deprecated functionality of an old version of some underlying service. Updating in-house parts to make it work with new versions is not made possible because “Phil knew how but Phil was laid off 10 years ago” or “the company who made it is out of business” or “we don’t have the money to do it” or “it works now, so why bother?”
-The servers are fine, up-to-date and secure, but the in-house service itself has glaring security issues that go unfixed due to above reasons.

And thus came along little Bobby Tables and was able to completely incapacitate his school district…

Generally a Linux installation is very good at keeping itself up-to-date and installing security patches automagically. Updating Docker containers is somewhat more involved, but can be easily automated with Watchtower.

Most of the services you use every day run on Linux servers. Even Microsoft uses Linux on their servers. And these services, not an average laptop, are the main targets of malicious actors.

The vast majority of behind-the-scenes infra that the end user never sees are open-source, even if the end-user part is proprietary. Eg. Facebook and Xwitter are proprietary, but run on open-source infrastructure like Docker, Kubernetes, Nginx etc.

Proprietary OS-s are workstation/office/home PC land. They have way more security issues due to crap coding whereas security problems with open-source server stuff are as a rule the fault of the admins misconfiguring services and not keeping their software up to date.

A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects.

—Robert A. Heinlein

family is everything, child need them

My favourite part is when the conservatives start talking about all children absolutely needing mother and father. Not just parents, not a parent, not a family; mother and father specifically. Yeah, sure, now what about the millions of single parents? Shall we start forcefully assigning a new spouse of opposite sex to them the day after their current spouse dies, divorces (if we keep that as an option, that is), runs away or whatever? All pregnant people who are not in a relationship are immediately married off to a random person of opposite sex? No opt-out. Because think of the children!

The very day Hamas attack happened I suspected that there’s no way Israel’s intelligence orgs, some of the best in the world, did not know about the plan for the attack way before it happened. This whole affair stinks to the high heavens.

Obviously somewhere along there were some hubs that weren’t obvious to the naked eye.

Probably the port on your laptop was on a hub built onto the mobo. If my understanding is right that’s how USB ports usually are connected: controller->hub->ports. If you open up device manager and go to USB controllers you’ll see several root hubs and hubs even when you have no external hubs or docks connected.

Navidrome just seems to be faster and more responsive. But the main reason of using both is that I just like to try things out and tinker. I also use Foobar2000, Kodi, MPC-HD, AIMP and other media players.

I actually bought just one new 6TB HDD and repurposed an older 3TB one as a redundancy drive for mirroring most critical data using a simple rsync cron job (no need for realtime mirroring of media files that are write-once), plus another old 1 TB drive just because. I haven’t run out of storage yet and I have automated download/sharing for OpenStreetMap and some Linux distros which takes up half a TB or so, but I plan on expanding the array using MergerFS and SnapRAID when the need arises.

The rest is just SMB shares, Navidrome, Jellyfin, DLNA and FTP. Remote access from outside my local network is done via Tailscale VPN.

I doubt 0.1…0.2 mm aluminium sheet metal is good for any scissors🙃