Test Equipment Anonymous

BU508A wrote: ↑Wed May 06, 2026 2:49 pm It is the current that kills, not the voltage.

It takes less than 150ma to kill you if it goes across your heart.

Part of the danger of getting shocks is the involuntary movement that follows. A shock from a spark plug can cause a hand to come in contact with dangerous moving parts. A shock from a CRT could cause an awkward fall onto something.

EC8010 wrote: ↑Wed May 06, 2026 2:17 pm Apparently, a shock can cause the release of chemicals into the blood stream that can kill up to 24 hours later, but prompt treatment can prevent this.

Any reference for that mechanism?

Sadly, no. I didn't query it when it came out as a company edict because it seemed like a sensible thing to do. But it would be nice to know why.

When I was a kid my Father was working on a portable TV and had the chassis pulled away from the CRT. That meant that the ground strap to the aquadag coating on CRT was charging up with no place to go. (Inner and outer coating are a giant capacitor) He powered up the TV and it had a perfect pix. I pointed my finger close to the face plate of the CRT mentioning how good the pix was and that capacitor discharged right through the front glass probably about 14kV in one spark into my finger. Hurt like hell.

I'll bet the verbal whipping your dad got from your mum hurt more.

BU508A wrote: ↑Wed May 06, 2026 2:49 pm It is the current that kills, not the voltage.

I guess it's time to link this video again: https://www.youtube.com/watch?v=BGD-oSwJv3E&t=10s

EC8010 wrote: ↑Wed May 06, 2026 4:59 pm I'll bet the verbal whipping your dad got from your mum hurt more.

I don't recall her saying anything.

I expect the words used weren't for your young ears.

tggzzz wrote: ↑Wed May 06, 2026 4:09 pm

EC8010 wrote: ↑Wed May 06, 2026 2:17 pm Apparently, a shock can cause the release of chemicals into the blood stream that can kill up to 24 hours later, but prompt treatment can prevent this.

Any reference for that mechanism?

Sounds like an old wive's tale, but death by electrocution doesn't present ready opportunities for ethical research.

I looked on the Internet and found nothing to support it, so you could well be right. Sadly, I also found British Medical Journal stuff that was littered with errors. I don't have much faith in medics. If you present with something obvious, they can often help significantly, but it has to be really obvious for them to get it right. I've twice been turned away with broken bones, with the medics telling me I was wasting their time.

You are certainly right about the ethical research bit. All of the limits associated with RCDs comes from work done during and before WWII.

EC8010 wrote: ↑Thu May 07, 2026 8:48 am I've twice been turned away with broken bones, with the medics telling me I was wasting their time.

Sometime it is necessary to insist, and mutter words like "I have a complex medical condition".

I managed the opposite. I walked on a painful foot and after a week I cycled 6 miles to the "nearby" A&E. Turned out I had a very obvious inch-long spiral fracture of a metatarsal. The doctors did nothing, with my agreement - since I was wearing hiking boots with a stiff sole, and that was as good as a cast! Naturally I cycled back home.

Went to a job interview in a suit with those boots. Got the job and enjoyed a few years at a zany company.

You are certainly right about the ethical research bit. All of the limits associated with RCDs comes from work done during and before WWII.

And presumably anecdotes thereafter.

I would have thought that a biochemical "24 hours later" pathway would be present in animals, and could therefore be ethically researched.

I suspect that the "release of chemicals into the bloodstream that can kill up to 24 hours later" is a "mistranslation". It seems more likely that damage might not be immediately apparent, but might manifest itself after a day. Such damage would be apparent through chemical markers. Hence an inversion of cause and effect.

My lack of clarity. I should have said that the chemicals are released following shock, but that they could take effect up to 24 hours after.

After being knocked off my bike, I cycled three miles home with hurt arm that second visit to hospital revealed was broken and needed an operation.

Adrenalin is released following trauma. I've noticed that after a near miss in a car, say someone pulled out and I had to brake sharply, I've felt an adrenalin rush. Since there's no real threat, it's a nuisance to be wound up like that. After a minute or two it wears off and everything seems strange. I haven't noticed it after electrical shocks.

I had a quick look on the WWW and I couldn't see anything about chemicals released after electrical shock, which were dangerous or needed management.

Zenith wrote: ↑Thu May 07, 2026 1:23 pm Adrenalin is released following trauma. I've noticed that after a near miss in a car, say someone pulled out and I had to brake sharply, I've felt an adrenalin rush. Since there's no real threat, it's a nuisance to be wound up like that. After a minute or two it wears off and everything seems strange. I haven't noticed it after electrical shocks.

After I fell and snapped my patellar tendon and the ambulance had arrived and assessed me (i.e. maybe 30 minutes), my arms/hands started shaking uncontrollably for a minute or two. That was the after-effect of adrenaline. I hadn't experienced that before; it was interesting to observe.

Quote:
"A web page that tells you what your browser gave away the moment you arrived. No login, no form, no permission. Most pages do this. None of them tell you."

Link:
https://sinceyouarrived.world/taken

That's basic stuff built into the fabric of http. Add cookies, and most of us can't be bothered to reduce them to the bare minimum, and other means of tracking and profiling you. There's a big business in harvesting and selling your profile for directed advertising and less legitimate uses.

If you feel strongly about it, there are ways of limiting it and enhancing security.

https://www.zdnet.com/article/how-to-st ... cking-you/

You can go further, for instance by having your own Domain Name Server, which avoids your ISP easily gathering data on your activities via DNS requests.

Not sure whether having your own DNS Server is a useful defense against

There are, of course, more intrusive mechanisms This is a rathole that is just as interesting as TE, and almost certainly deeper than TE. I have avoided falling into this hole, up until now at least.

I've just read about the Tails distribution intended for journalists and activists. https://tails.net/index.en.html That also includes a decent discussion of its failure mechanisms[1], and of other software it contains. For example, applications that strip identifying metadata from files you want to send to someone, and applications that sanitise documents you receive from untrustworthy sources.
https://gitlab.com/metadatacleaner/metadatacleaner
https://www2.tails.net/doc/persistent_s ... ex.en.html

Can't be bothered with Tails, but I wish I could run qubes (A Reasonably Secure Operating System) https://doc.qubes-os.org/en/latest/intr ... s-qubes-os on my machine since it would help avoid standard commercial (mal)practices. I can't because I have an nvidia graphics card, and the drivers for that are an impenetrable blob. One well-known engineer, Phil Hobbs, runs it for his engineering activities.

[1] amateurs/beginners think about how things work; professional think about how things fail.

Tails is another specialised Debian distro with selected software. So nothing you can't add to some other distro yourself, but obviously more convenient.

Qubes OS looks interesting. I haven't looked into nVidia problems, but it occurs that since it runs a string of virtual machines, it's going to soak up RAM and compute power, so 16GB, and better 32GB, of RAM to be on the safe side, and a fairly up-to-date CPU, say released in the last 7 or 8 years. It implies that to set it up, get it running smoothly and then commit to full time use, you have such a machine hanging round doing nothing else.

My experiments along these lines involve a thin client, which uses < 10W, to run various services, such as a Pi Hole and torrent client. A lot of these things will run on a Raspberry Pi, but the ones I have only have a USB SSD connection, which isn't great. For instance, most USB SSD adapters don't support Trim.

The security is in the details, and there are a lot of details! Hence I regard tails as being more than just another distribution.

Joanna Rutkowska's blog post is useful https://blog.invisiblethings.org/2012/0 ... -from.html

The qubes apps VMs aren't like the VirtualBox/VMWare-level VMs; the qubes VMs are much more lightweight. In particular they all presume they are running the same root linux (and hence files), which enables a lot of "read-only" sharing. That means adding another app VM doesn't mean that the RAM requirements are doubled. Additionally memory can be "rebalanced" between qubes. No, I don't know the details.

Have a look at https://doc.qubes-os.org/en/latest/user ... qubes.html to see how "Carol" has a separate qube VM for each website she wants to isolate from others. I also quite like the concept of disposables (VMs) https://doc.qubes-os.org/en/latest/user ... ables.html You can take that further so that each each email attachment gets opened in its own single-use disposable qube. https://doc.qubes-os.org/en/latest/intr ... n/faq.html

I don't think the compute power will vary very much as app VMs are added; in general the Xen hypervisor stays out of the way almost all of the time. I don't know how much compute power is required to communicate between qubes VMs. I doubt it would incur much more overhead than a microkernel operating system.

Besides, compute power is more or less free nowadays. Cache and memory bandwidth less so, but that the same in any OS.

Wish I could try qubes. Maybe I'll buy another graphics card

There is a FOSS nVidia driver for those with an objection to closed source blobs. I don't use it because it's noticeably slower than the proprietary version, but it works. You can swap to it and back using a GUI app. Other than that, if you are not into gaming, or only a little light gaming, video cards, which are not much more than video adapters, and so are rightly sneered at by self-respecting gamers, are available on ebay for not much. Then you could always find a used APU, such as an AMD 3400G, on ebay.

I looked at the hardware requirements for Qubes OS and it was 6GB RAM and a 64bit CPU, but it was quick to point out this was the bare minimum and more was highly advised. The SSD cpacity required isn't excessive. Browsers are the killer. In my experience Firefox with a lot of tabs open for a few hours, will cause heavy swapping with 8GB of RAM and can cause the PC to lock up, or run so slow that it may as well be locked up. Because of the LLM boom DDR5 prices have become prohibitive and DDR4, which is going out of production, is also much more expensive than it was.

I don't know what the second hand scene is like for used PCs and laptops. I believe there can be pitfalls.

Thinking about it, I suppose I could run Qubes OS as a virtual machine on this PC, (32GB of RAM and an AMD 5600G APU). A string of virtual machines running in a virtual machine. It all seems like rather a lot of trouble and whatever could go wrong?

I have a third serious PC with a 15 year old 4 core 4 thread AMD Phenom II, which I keep for experiments and to run those MS applications I sometimes use. I'm not sure it supports virtual machines and its 4GB of DDR3 RAM would definitely need upgrading. It only has two RAM slots. OK for interesting flavours of Linux etc, but probably not for this.

Let us know if you make any progress with Qubes OS. From my POV, it's intriguing, but a bit of a pain to try.

Zenith wrote: ↑Mon May 11, 2026 6:20 pm Let us know if you make any progress with Qubes OS. From my POV, it's intriguing, but a bit of a pain to try.

I doubt I will make any progress; I share that PoV.

It's just occurred. It could be done by loading Qubes OS onto another SSD. It wouldn't need anything fancy to try it. The BIOS could be used to select the boot device and it would be completely separate from the main system I rely on to work. 256GB SATA SSDs used to be about £25 for an OK, but not special one. They're more like £50 now. I have a couple of spares. It could even use an HDD. In your case you'd also have to do something about the video card, if the nVidia driver is a show stopper.

OTOH, I have a number of more pressing projects.