soneil

My previous contract (Ireland) was 3 months - it was reciprocal, so they had to give me 3 months notice to quit, and I had to give them 3 months notice to quit.

Due to the position "garden leave" was the most typical way to spend those 3 months, so it wasn't much of a real-world concern.

I find it difficult to side with Delta when they're trying to shirk their legal obligations to their customers.

There's a comment in the video that this is david vs goliath, and they give david some leverage. I think that's the perfect use for high-court bailiffs. The other way around, maybe not so much.

(Ironically - obeying court judgements is not without control of the airline, so if this caused delays, those customers could claim compensation also.)

There's a strong chance Delta don't have much legal/fiscal presence in the UK. Probably also why they didn't feel a pressing need to pay fines in the UK.

There's a much higher chance that Twitter have a legal and financial presence in Ireland, especially being that it is/was their EMEA HQ.

Appears it didn't progress after that either - https://datatracker.ietf.org/doc/draft-looney-tcpm-64-bit-se...

The growth seems entirely in line with the curve. I'm not sure it needs much explaining, let alone guesses.

It looks like the static is tiled. So it's random (enough) within a tile, but you can perceive the tiling.

To be fair, that's pretty much the intended usage. The SSN is supposed to enable them to figure out exactly which John Smith they're looking at.

It's a serial number, not a shared secret. It sounds like your college treated it as such.

The real problem with SSN is the prevalence of unintended usage.

You can use both at the same time, but you can only use two cores. eg 2xARM, 2xRV, or one of each.

yes - two votes allows you to detect a disagreement, three votes allows two votes to win.

It's a bit of both - you do have to pull out your wallet, but there's also an approval process. Just because you can buy a gTLD, doesn't mean you can buy .con

.local being used for mDNS while Microsoft were using it in AD examples/documentation is another good example.

.internal is just admitting there's only so many times we can repeat the same mistake before we start to look silly.

I'm pretty sure this is why everything we got in the first 48 hours from CS was stressing that the issue was with a "channel file" (threat definitions, content updates, etc).

Their staged update process is for the falcon driver itself. It is not for the "channel files".

As I understand it, the driver itself is understood to be a risk, and they provide facility for an N, N-1, N-2 staged deployment to mitigate this risk.

As I understand it, channel files were not identified as a risk, and were never subject to this staged deployment.

The "sell" was that you could be running a trusted driver at N-2, but still have 0day protection from up-to-date channel files. And CS's initial feedback that the issue was not with the driver itself was CYA that they hadn't been misleading customers using such staged deployments.

So straight up I have to admit that this isn't my wheelhouse - I support a bunch of developers who seem to enjoy breaking things. Safety-critical or life-critical just isn't my thing. If breaking stuff is half the fun, you probably shouldn't be in medicine ;)

Say you have one server that houses all your patient data, and 1000 workstations that access it. I think it's safe to assume you'd treat that one server as your "crown jewels". You want it to be triple-redundant, you want it to be on battery, generator, a very conservative lifecycle management, replicas in different fire zones, immutable backups, etc etc.

Your thousand desktops .. meh. This is where you want your endpoint protection, this is where you're worried about data egress, etc. They still need to be controlled because they have access to the patient data. But you're not so worried about resilience. If a workstation goes bang, you just go out and image it.

I'd consider this a fairly typical way to evaluate risk and threat.

"I felt a great disturbance in the Force, as if millions of voices suddenly cried out in terror and were suddenly silenced. I fear something terrible has happened."

So on Friday, those thousand workstations simultaneously turned blue. Our hypothetical threat model so far has treated workstations as disposable, replaceable, but didn't consider workstations in their entirety. And once we lose the entirety, all our "crown jewels" are safe on our triple-redundant servers, but there's no way to access them. And the resulting "stop work" is a risk to any patient who really needed that work done today.

Now as I said, this isn't my area at all, I'm spit-balling here, but this is how I understand the fallout from this. An analogy is that we put more effort into protecting the president than the man on the street - but if you wake up one morning and the general population has disappeared, the impact is bigger than losing the president.

> Whoever thought up the great idea to allow auto-update-able kernel modules

What's made this whole thing so "interesting" is that the whole point of these "channel files" was to decouple the risk from updating the kernel driver.

Accepted best practice for this product has been to stagger rollout of the kernel driver, so a pilot group gets the current release, the herd get n-1, and sensitive machines get n-2. The product provides for this, and most sites either use it, or admit they should.

So when your pilot group start bluescreening with "DRIVER OVERRAN STACK BUFFER" (actual example from last year), it's caught (by the customer, still) and triaged before it reaches n-1, let alone n-2 & front page of The Times.

But the whole 'sell' of the product is that they get 0-day definitions. So endpoints running the relatively trusted n-2 release still get the same protection against active threats. n-2 have a stable driver running today's "channel data".

I'm not clear if Friday's "channel file" is the issue in itself, or whether it triggered a less-explored code path in the kernel driver - but the result is the same. The best practice of staggering the kernel driver releases, didn't save us from a logic bomb in the "channel file".

I just think the distinction is interesting because following accepted best practices, vendor recommendations, and conservative deployment recommendations did not protect from this. It's not the customers that were yolo'ing this.

This is going to sound incredibly cynical, but I'm convinced the primary reason for Smart TVs to exist is to sell TVs.

My current TV is 17 years old. It's 720p, plasma, and yes I want to replace it. But functionally it's as good as the day I bought it. I've used it with appletvs since the first generation - if you're not into apple, there's android and amazon boxes and sticks too.

When my appletv goes EOL, I'll buy another $99 appletv - and they're usually supported for about 5 years.

I do fancy upgrading to 4k and all that jazz, but every time I go looking at my options - they're all "smart". They're rarely supported for 5 years, let alone 17. Obviously yes you're going to airgap it once it stops being supported in 2-3 years, but having to click through one OS to get to my STB's OS sounds like such an ugly kludge.

So back to my original premise - I can't help feeling that annoying me into upgrading in 3-5 years rather than clinging onto it for 15-20, is the entire point. The primary reason I want to upgrade today is to go from 720p to 4k. In 17 years I'll be touching 60, and probably won't be very fussed about 8k. If I could buy a current version of my venerable old plasma today, it'd likely be the last TV I buy. And that's not what they want. They've seen the lifecycle on phones and they want in.

(rant aside: when I got my current set, I naively thought CableCard was going to evolve into pluggable smarts. So you could buy a brain from your cable company, or apple, or msn, or tivo, and slot it into your TV. And when Real10 needed more grunt than your card had, you'd get a new card. Instead we ended up with disposable TVs.)

I'm not really sure cloud has much blame here either.

Imagine it's 1998 and Norton push a new definition file that makes NAV think kernel32 is a virus. The only real difference today is that always-on means we all get the update together, instead of waiting for mum to get off the phone this evening.

We got an email this morning telling us none of our usual airlines could take bookings right now. That wouldn't have been much different in 1998, airline bookings have been centralised for my entire lifetime.

It does seem to somewhat skew the aim of the site. Account for all the black kids riding rings, account for those near depth, and then watch to see which changes. I got bored before I found any variation on the theme.

I did find a couple where you could tell which kid was going to be in trouble long before it'd let you click them, which I found interesting.

It's part of PE (phys.ed) here. It was just like any other class - one morning a week for one school year, you're dragged off down the muni pool.

New words are invented all the time, but girl, boy, dog you'd expect to be ancient concepts with ancient etymologies. 'noob' has an obvious etymology in new-neue-niwe. The idea that 'dog' doesn't blows my mind - especially since the guttural, monosyllabic utterances are often the oldest.

This is how I store my one-off, unicorn, proprietary, or just "retro enough that I don't need 50 of them" cables. Fold the cable around the width of my palm, then stick it in a TP tube, stored end-up in a drawer with both connectors facing me.

Very searchable, very retrievable, very cheap/available, and stops them coalescing into a lovecraftian tubmleweed.

Within the EU at least, "sin taxes" work well for this because carrying them across the border for personal consumption is the only legal workaround. For most other things, if they're notably cheaper in another EU country, I can just order it from that EU country.

For example, I'm in Ireland (relatively HCOL), and I regularly visit Slovakia (relatively LCOL). When I used to smoke, it was 3.30/pack in Slovakia and 12.50/pack in Ireland. I could legally bring 40 packs back with me, so €368 to be saved - back then that would not only cover flights, but also a couple of nights in a half-decent hotel. (sunken-cost assumed, obviously).

But to compare, for example, an m3pro macbook pro, the difference is €100 - at which point you're really only juggling the difference between 23% and 20% VAT rates.

Outside of europe the gloves come off. Brazil, for example, has famously high import duties - so smuggling the same macbook to Brazil would have a much greater pay-off.

The limit for duty-free has existed for a long time (in 2002 I bought a a 5x10x20 pack on my way out of the country, and was warned I wouldn't be able to bring it back. This is quite a distinct memory as I remember quite proudly telling them I wasn't coming back!)

I believe what's been lost is the "booze cruise" intra-EU option (where tax has been paid elsewhere in the EU) where the limit is "for personal use", an incredibly vague limit where the onus is on you to convince customs that you have no intent to sell.

It's worth pointing out that the personal allowance is usually for personal use, intent to sell does not come under personal allowance.

Usually this is near-impossible to enforce, but attempts to organise/aggregate these (which I initially thought this site was going to attempt) do raise your odds.

The second-last paragraph seems to cover this. In short, no - but it seems reasonable to believe this could be a solid stepping-stone on the path from here to there.

$40 shipping (to Europe) on a $45 paperback was a tough sell, but you're right - his SNMP "Networknomicon" was funnier than any book on the topic has any right to be. I think I read it in two sittings.

I think this itself is playing off an older deeznuts/ligma gag.

setup: There's no countries in Africa starting with the letter K

fallguy: Kenya?

punchline: kenya lick ...

My biggest issue with this isn't that it's pretty much parroting a knock-knock joke as a factual account. One is that it actually has no idea how to reach a factual answer - it's not digesting source data and filtering it to match the request, it's just restating the first search result as fact. My other problem is that it only seems to care about the first search result. It can't summarise the first 20, 50, 100 results and find what's consensus, what's disputed, etc.

I realise we don't usually expect that of generative AI, but this is generative AI which should be hooked up to google's search results, but instead is 2024's "I'm feeling lucky".

I suspect a lot of it isn't just this feature in itself, but the overall fatigue of having it shoe-horned in everywhere. I recently had to replace the Logitech software for our mice with the (hidden in a kbase) offline version because they'd added AI to what's essentially a mouse driver - two weeks later it pops up in my terminal.

I think for most people, it's not any one mole that's the problem, it's the overall game of whac-a-mole.

That's really not far off Scotland's "responsible access" (the proper term for the 'right to roam'). It's just more fleshed out.

The Outdoor Access Code contains such gems as "You only have access rights if you exercise them responsibly", which I love - it's the closest I've ever seen to "don't be a dick" being codified. But also such details as in crop fields you only have access to the margins, fields with animals, bringing dogs, etc.

But they do have "land on which there is a house [..] and sufficient adjacent land to enable those living there to have reasonable measures of privacy and to ensure that their enjoyment of the house [..] is not unreasonably disturbed". Which isn't "x metres", but is again wonderfully close to "don't be a dick".

They don't have set times for camping, however.

I think it's an interesting quirk that we don't use the same terms for what are essentially the same concepts.

Switching is $verbing at layer 2, IP routing is $verbing at layer 3, NAT is $verbing at layer 4, and reverse proxying is $verbing at layer 5.

I'd argue that 'route' is the right value for $verb here.

It doesn't sound like they're trying to kill them, since they're moving commercial usage to a subscription license.

I'm not sure killing them actually makes a lot of sense - the products apparently share a lot of code with esxi, so it's two products for the R&D of one.