Posts by rgjnk 239 publicly visible posts • joined 16 Jan 2024
Done in weeks can be done, has been done, just isn't cheap & requires shedding a lot of process & paperwork.
Sea King AEW with Searchwater was done in 11 weeks, other stuff has also been turned around quickly.
It's not the engineering that takes the time, it's everything else.
Given how they're all tying themselves together in these stupid circular deals with little new/real money behind it, it's not going to take much of a problem in one company to screw a whole stack of players.
And best of all so many of them have got themselves into a position of the AI bubble being the vast majority of their business.
If you're relying on being able to read the code to trust it you aren't trying hard enough.
Assume you can't trust it regardless and wrap it up appropriately.
I have been down in the weeds of cloud and hypervisor code which is great if you're looking for something specific (or want to fiddle) but it doesn't mean I can *trust* it, whatever I mean by trust. Many eyes are not a guarantee and there's no way to independently validate the whole stack yourself either.
I might have invested (under protest) in the on-prem Azure solutions if they hadn't always felt like an unloved step-child that Microsoft would rather didn't exist and they didnt really want to sell to you.
Changing road maps, poor support, poor updates, expensive, hard work, and historically too close to the mothership.
They want you on their actual cloud and that's pretty much it, so you have to really want something specific from MS for the other Azures to look tempting.
If you want on-prem cloud or HCI solutions it's easier to look elsewhere.
Been there and done it with OEM level resources and better documentation & familiarity than your average end user nation state would get. It's a very very specialist task, even within a specialist field (in part because it's not usually necessary) and doing it on the actual platform is asking for disaster.
And that's assuming anyone jumping in would have a clue what they were looking at, avionics isn’t based on the standard BSPs or bootloaders or schedulers or open source. They likely wouldn't even recognise how the processor was configured compared to the standard off the shelf initialisation 99.9% of people use.
It's one thing to fiddle on a rehost and have it work 'well enough' after patching, doing it onto the actual avionics will likely work just well enough to break it, or potentially just trip over some hostile anti reverse engineering features that everyone just loves to put into their special military systems that run a risk of being captured & analysed by nation-state level resources. It's not a PlayStation or a phone.
And you'd be very very brave to trust an unofficially patched aircraft software load. It takes long enough to get a proper one out the door, especially if you actually want it mostly bug free.
The Dutch can make all the claims they like but it just isn't that easy. And they haven't even got a relevant domestic OEM or supplier to help out.
As for a remote kill switch, seems deeply unlikely, too much risk. Something that nobbles people operating outside the official supply chain after a while? Quite possible, and that can be very sneaky, under the pretense of ensuring the users are operating correctly with OEM support.
The product and their responses sounds like they're a long way down the rabbit hole of hacking stuff around using vibe coding.
None of this sounds at all like structured product design with vaguely intelligent people in the loop.
The rapidly shifting role of 'verbose' is certainly a clue.
6 years to turn a profit? Everyone knows the numbers are fantasy, they stretched it out to make the finances look better.
If you're running the hardware hard enough to generate the necessary returns it isn't going to last that long. Assuming anyone still wants the access to that spec towards the tail end of that and it's still worth lending it the rackspace.
What is he trying to achieve with a me-too AI effort? What is the end goal here?
It looks like 'AI -> ??? -> Profit' but without any obvious clue about what the actual product will be, let alone how it will generate any income.
And that's before we get into plowing $$$$$ into infrastructure apparently just for the sake of scale rather than to satisfy demand that they can't even define the shape of yet.
Wouldn't be a surprise for someone to have made a clumsy effort to remove their/their mates details from the dump, while forgetting that an admin with the original can run a quick compare to spot the difference.
Plus, who's dumb enough to put any meaningful or traceable data into this sort of forum in the first place?
Thinking two steps ahead can be a challenge for many.
It's not that long ago that people really started to become keen on securing everything with signing and certificates, and plenty of stuff just chucked in certificates with 10 or 15 or whatever years of expiry because that was way off in the future.
Just a shame when the future actually arrives and the certificate expires and stuff breaks.
Some things don't care about expiry but more than enough do.
First off - there is some exceedingly rapid obsolescence in the CUDA versions vs the hardware anyway, and that's before we get into how different generations of the GPUs are tailored for specific load types and not wildly useful for other/newer tasks - you can see this from the retained value/disposal at knockdown prices of the older generation hardware.
Secondly you're assuming something else will (can) step into the gap. There is/was a lot of kit floating around from the earlier metaverse & game streaming bubble and no-one wants it, even though in theory it's nice cheap gaming capable gear. Partially as the market to run it commercially doesn't exist, but mostly because the hyperscalers customised it and there isn't any driver support anyone can get. It's orphaned.
All that hardware will be rendered worthless, it's just a matter of when. This is true of all hardware. It won't generate a new market just by existing.
Will the tech survive and evolve? Probably. Doesn't mean the existing major player will have much to do with it, in the same way previous giants and the gear they produced rapidly faded under market changes. Nvidia as it is now is a recent creation and all that can fade again to its pre-AI/crypto state.
Have they? Really?
So far all I've seen is the old services drop out and some apparent hardware retirement/mast consolidation if the worsening signal strength on 4G and the rest is a guide.
Certainly yet to see any improvements, I'd be happy if I just got the level of coverage I used to see.
They would be far from the first European organisation with strong security requirements to have their own private fully sovereign cloud. It has been done repeatedly, and at scale.
Google seem to usually provide most of the tech because they're quite happy to sell you the full stack for you to run independently. Usually in conjunction with someone else wrapping the whole thing in whatever security you want.
Other options exist too but for true sovereign cloud Google always seem to win these days.
I suspect their bigger problem is working out how to shift their existing infrastructure to any sort of cloud environment as they probably have all sorts of specialist stuff that works nicely on-prem and will be a pain to shift.
They promise all these wonderful things, but I've heard plenty of pisstaking about Palantir just reselling the same old solution to every customer with lots of promises of how it fixes everything. Then doesn't.
Good while it lasts (like their stock price) but they aren't changing the world.
Also it's run by arseholes who manage to somehow make the competition all seem like particularly well grounded saints. Now that's a real achievement.
Google seem quite capable of utterly breaking their AI with updates. Like (for example) from one week to the next moving it from having lots of context/conversation memory to having total amnesia between prompts.
Lots of jokers in the AI realm pretending they know what they're doing and Google have joined the rest in hiring them.
So if they're proposing AI to fix AI I have some doubts...
Slide & push is the sort of thing you could do just by having the thing loose in your pocket, it's hardly a great protection against deleting by accident.
I've got some nice secure USB devices but I'm not too keen on ones that lose the data I've spent extra money on protecting.
All sounds a bit like a toy overall.
I get the security points they were trying to make, but I'm not sure anywhere under testing the data access or the rate limits or anything else required pulling as much of the data at they did.
They pulled *all* the records when pulling 0.1% would still have been more than enough to achieve exactly the same.
What they did just feels unnecessary.
You'd almost think there weren't already solutions available off the shelf for partitioning & isolation, with aircraft certification, that could run on commodity embedded processors. They even have all the useful things like proper real-time support and IO drivers and audio and graphics and anything else you'd ever need.
But no, let's not look outside their comfort zone and instead let's try to reinvent something else into something it was never designed to be.
They're starting from the wrong place and doing it the wrong way because they don't know better and haven't learnt the hard lessons yet.
Reading the full statement it really reads like they want Gateway API to become the 'thing', even though it's not Ingress and doesnt do the same job.
As ever there might be a bit of a disconnect between what the project wants and what users want. And the use cases certain people have (which are all that apparently matter) vs everyone else.
Also funny that they sell this as a security move to protect the ecosystem yet go out of their way to say it will keep on working - most terminal security issues involve a very rapid halt not a few months of best-effort maintenance & the ability to carry on.
Also the total lack of migration planning is a bit of a hole...
So a widely used, popular baseline option is being retired because they basically can't be arsed to support it?
There's nothing insurmountable about any issues (if it was that bad no one would use it), it seems more that it's under resourced and no-one wants to put the trivial effort in to sort that.
Spend less time on your onanistic committees and more on making sure key components actually have maintainers. All this stuff is a business creating revenues in the billions and allowing many to indulge their merry talking shops, yet no one seems to be bothered to do the actual work because it's not the latest shiny.
Maybe someone could take the time off from felating themselves with a Rust version of sudo and put it into code for something more useful?
"While Tencent’s capital expenditure fell, its revenue and profit rose."
In the rush to join the bubble, maybe a wiser mind realised that 'investing' in incredibly rapidly obsolete infrastructure that never generates enough revenue to cover the install & operating cost is maybe the better move?
Supply constraints is a gentler way of selling it than saying "we're staying out of this", meanwhile others plough $billions into kit that will never generate a return.
"Sustaining open-source work is not a one-size-fits-all challenge."
Unlike, say, the endless promotion of a certain language as the one-size-fits-all solution to every problem.
Also - "keep the language stable and evolving"; well, it's either one or the other isn't it? Seems to be leaning much more to 'evolving' at the moment.
I suspect many would be happy if it did burn out as it's mostly about fervent promotion into unnecessary spaces right now than genuinely being better than the many many alternatives or the status quo. The big party trick isn't exactly unique or special.
It's almost comical if the examples are anything close to real, it's such basic old stuff!
Chinese knocks of products? Happens to anything popular even without spys, from widgets to specialised test tools.
People sticking USB where they shouldn't? Nothing new - users should educated, and the possibility of abuse blocked. Especially when off site. Lock it all down!
People on tours looking or taking what they shouldn't? Welcome to something that has happened forever. If it's a risk then either don't do tours, be selective of who can come in, or hide the special stuff away. Some people are just too stupid or trusting and get ripped off.
People asking questions? Welcome to basic intelligence work since the dawn of time. Again, some people are just too stupid or trusting and shouldn't be given information they could leak.
It's a worry just how many naive idiots I have to share air with, every time an obvious test phish comes around a large percentage fall for it and hand over their credentials - if they fall for those then an actual sophisticated campaign will snare plenty. And this is in an environment where they're meant to be smart and experienced enough to already understand these things *and* have been specifically trained to handle the threats.
And still the morons fall for it.
.
'He said the technology does "not monitor innocent shoppers. It does not store your data."'
Well the first is obviously untrue - it monitors everyone because that's the way these things work. 'Innocent' is an output after you've been monitored. The only thing I'm slightly unclear on is whether it's a gate type system on entry/exit or store wide monitoring.
As for not storing your data - really? The data goes in and I only have their word about what elements it might retain and for how long. It's obviously capable of storing some sort of data for some period otherwise it wouldn't have anything for comparison or for reporting.
I can understand some of the motives for the CEO but the attitude in the statements is a bit too confrontational; when you're in that zone all sorts of things become apparently justifiable when they wouldn't under more considered thought.
Lots of (most?) commercial property has tenants who are fairly sticky for various reasons.
And yet a look at the commercial property market over the years suggests that doesn't necessarily make it a great investment - things change, sometimes rapidly, the tenants evaporate and you're stuck with a chunk of vacant property that's costing you money to own and you can't easily shift. If you're lucky it's generic or repurposable, if not...
Shopping centres/malls, office buildings, business parks, high street shops - all have been hot, then not.
The collapse of this latest hype is entirely predictable yet many will still get caught when it happens.
I have deliberately run proper calibrated tests to vibrate a disk to the point of error (they were going into kit in a harsh and moving environment) and it takes a reasonable amount of proper vibration to get a drive to persistently error. I never managed to outright kill one.
It also takes a significant amount of energy to vibrate a big heavy rack - the actuator we mounted my test rig on was *big*.
A fixed in place old telecoms rack full of servers is a very heavy solid lump of metal, and what would inevitably have been a battery drill/driver is not going to meaningfully shake it - even a high impact SDS wouldn't. More likely it would be the operator doing the shaking as the much lighter wobbly fleshy part of the equation.
And having actually drilled holes into a rack with a battery drill 1) you wouldn't accidentally have it on hammer for more than a fraction of a second, and certainly not hard, and b) you'd soon work out it was a lot of noisy effort & find a lazier approach.
So not to cast doubt but based on science and hands-on experience if someone managed to kill their systems it wasn't via a drill.
Plus - as a final killer - racks are cheap. With all these purchase orders where was the plan to adapt or replace the rack? Anything I've ever seen which was incompatible wasn't ever going to be fixed with a drill.
As a (slightly) informed opinion of how it will be:
- They've got a lot of integration and innovation type stuff promised on the basis of 'how hard can it be?' with no prospect of it actually happening for various technical reasons. The people directly involved don't know what they're asking for and/or have no ability to deliver it. Wiser minds will take funding to deliver the external technical reports about how it either can't be done, or can be done if given £££££.
- A lot of chancers will pile in offering various solutions, especially AI ones, but their abilities will mostly end at talking a good game rather than actually delivering. Shiny demos will be provided that have nothing underneath beyond smoke & mirrors.
- Real funding won't match the promises and, while significant, will be inadequate to deliver what's promised. Early days will be relatively cash rich, rapidly heading towards a shoestring setup a few months in.
- A management structure and a lot of assigned staff *will* happen as bodies are easy to find even if qualified ones delivering useful output is a bit trickier. Usual well known top heavy organisation patterns will be applied, and metrics achieved that somehow don't involve anything real being delivered. This may start out as a big Potemkin setup to look impressive, that rapidly shrinks as interest & funds dissipate.
This is not the first or last time someone will be setting up something (nominally) technical and innovative using taxpayer funds, new staff, a shiny facility and zero prospect of achieving the headline goals. The most anyone can hope for is to get close enough that some of the funds flow their way while it lasts.
You can hype your research by trying to act like a statistical model can be anthropomorphised, but the most that's happening there isn't deliberate sabotage, it's just the usual buggy output.
Other people have done hype where they implied their model had acted in some clever self aware way to avert its own shutdown but what they'd actually done was use a prompt to explicitly create that as the desired output.
We're definitely in the 'outright charlatan' stage of the hype bubble.
Wasn't/isn't there a related problem in the US with Doordash(?), pre delivery tipping, and the effect this had on the (non)service people were getting?
All these businesses with the same core model have a similar level of ethical/quality management, attract a similar level of workforce, and all descend into a similar level of mess in the end.
So was this a proposal to turn Local/HCI into something closer to Hub or just another round of confusing partial product that was sort-of Azure but not really enough to be actually useful?
I've look repeatedly at spending on their on-prem/hosted solutions (in fact have spent) and it's a real mess, especially when you get into costs and lifespan with their ever shifting ideas of what's available & stuff that dies of neglect if not outright killed.
The only constant is their desire to drag you into the morass of Azure-proper.
I'd steer well clear if it wasn't for a couple of things that force it to be used; it's not that their stuff is technically bad, more that (like a couple of other vendors) they're deliberately hostile to the customer & mess you constantly around for their gain, instead of it being a nice simple exchange of product for cash.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2026
