Posts by cyberdemon 3488 publicly visible posts • joined 26 Jan 2010
Depends on the project.
I was at the Embedded Open Source Summit EOSS2023 in Prague, specifically for the Zephyr developer summit. The demographics there are pretty healthy, a lot of young developers there, maybe because it's a fairly new and fast-moving project.
That said, there is a very high bar to getting code accepted to Zephyr, and for good reason. But the other devs are very helpful with pull requests to get it up to the required standard.
Well, if you as a client had paid for legal representation by a qualified lawyer, and you did not in fact get legal representation by a qualified lawyer, but instead were given generative slop from some bullshit machine, then it would be fraud. But here, it is the judge who has been presented with slop, so it is contempt.
I think it's pretty rare that two candidates would be exactly equal in their abilities. If it isn't rare, then it suggests that the hiring decision is not sufficiently detailed / granular e.g. rating everyone on a scale of integers between 1 and 5.
It's rarer still for those candidates to be of different genders, given the rarity of female applicants to engineering roles in the first place
So you are right, in an infinitesimal proportion of cases.
But, for the sake of sanity, I would just improve the detail of the hiring process to make sure that these cases don't happen, i.e. find out who is truly the better candidate, and hire them.
The whole idea of selling bulk access to HIBP seems to defeat its purpose.. The purpose is for end users to improve their security by changing credentials for accounts that have suffered data breaches. Whereas bulk access would allow a third party (perhaps a cybercriminal themselves) to fish for targets who might have forgotten to change their credentials, and to either target them with scams, or indeed go and purchase those leaked credentials on the dark web and see if they still work.
Meh.
When I got my Fairphone4 and installed /e/OS on it, I was looking for a de-googled, cloud-free phone with a micro-SD card for voluminous local storage, that I could back up myself via Syncthing.
Naturally, I didn't subscribe to a Murena account or use any of their cloud services, what would be the point? Someone else's computer etc.
So, I think the mistake was to bother offering these "services" in the first place. I don't want them, I don't know why anyone else would want them, and I don't know why they were offered in the first place.
Maybe, anyone who makes a mobile phone OS that gains any sort of traction soon gets a visit from some men in suits and dark glasses, who tell them that they must offer some cloud storage services.
Well, the reason for the backlog is essentially because the DNOs (Distribution Network Operators) and TSO (Transmission System Operator) are receiving requests faster than they can deal with them.
That is partly due to duplicate submissions and 'zombie' projects that were never intended to be built, but objections to "pylons" and substations from NIMBYs (and so-called NOBYs or BANANAs, who say "Build Absolutely Nothing Anywhere Near Anything") do obviously reduce the ability to clear the backlog, so it is logical to say they are partly responsible for it. And as I have said before, the NOBY anti-pylon argument is misinformed.
(That said, I think the NOBY brigade are right about Solar - The place for solar panels is on roofs and the odd sheep field, not high-grade arable farmland)
But, we really do need more overhead pylons and substations, net zero or no net zero. The grid is dangerously unstable without them, and there will be deaths if we have unexpected blackouts. Imagine your granny going upstairs to the loo for example when the lights go out. A certain percentage of the elderly/infirm would fall or otherwise injure themselves, as well as road accidents when the street/traffic/building lights suddenly go off, it becomes very dark indeed
Well, the "misleading clickbait" that I was commenting on is the "meta:og-title" which appears only when you share a link to the article (or read the HTML source, line 11 of view-source:https://www.theregister.com/2025/02/10/london_has_400_gw_of/), which says that the 400GW is all Datacentres, which OFGEM does NOT say, nor does the actual article or its title.
But 400GW just for Datacentres in London?
The articles's "OpenGraph meta-title" (the text that you see when sharing the article on something like Slack or WhatsApp, seen in the HTML headers) says "London has 400 GW of grid requests for datacenter builds" - whereas the actual title is (or has been changed to?) "London has 400 GW of grid requests holding up datacenter builds" which could mean that there is 400GW in the queue in total, including solar farms, gas/diesel generators, EV chargers, housing developments, etc. and all this is holding up the approval of the er, much-needed datacentres.
But then again, the article says "These figures are drawn from the EMEA Datacenter Market Update for H2 2024" so it might be talking about just datacentres, but it's one of those evil Javascript PDF "flipbooks" so I can't be bothered to read it.
Edit, OK I relented and read the ghastly flipbook - London has 265MW DCs under construction and a further 1260MW planned, so more like 1.6GW, not 400. So the reg "meta-title" is misleading clickbait.
+1 for turbine-electric hybrids
The turbine cars of the 50s used mechanical gearing to go from high RPM (100k+) to wheel speed and it was stupid. Gyroscopic loads and supersonic surface interactions caused bearings and gears to fail. But there have been significant advances in microturbines, high speed bearings and high speed low-inertia electric generators since, and an all-electric (series) powertrain means that the turbine runs at its optimal speed, and has no gears to wear out.
They are much lighter per kW than ICEs and have a similar if not better overall efficiency. It would solve a lot of problems with the 'range extended EV' concept.
I don't know about Texas, but in the UK, we have a lot of small "reserve" power stations which are essentially farms of containerised Diesel gensets. They get around the rules on carbon accounting, because the individual generators are below 2MWt - (but they have a few dozen of them at a time colocated on the same site..). These "Diesel farms" got a lot of press in 2015 but for some reason they have been forgotten about now. But whenever the grid is struggling, they make a massive profit on the UK's Balancing Mechanism, selling at up to 50 times wholesale price to the grid.
I have also heard of datacentres overprovisioning their "backup" generators by just over double, so that they can not only be paid to disconnect from the grid, but they can function as one of these Diesel farms at the same time.
However bad coal or gas is, diesel-powered electricity is worse - there are no electrostatic exhaust scrubbers, and their thermal efficiency is worse than an open-cycle gas plant.
What we needed was nuclear - but the industry has been regulated to death. Apparently a lot of that crippling regulation and anti-nuclear propaganda of the day came from the Rockefeller foundation, who had a vested interest in Oil.
At the rate we're going, a wad of £5 notes will cost less than a lump of wood or coal for the fire.. Unfortunately they are plastic, so they don't burn as well as they used to.
But, blackouts are a serious possibility now, never mind with Ed Milliwatt's barmy plan for 95% renewables.
I only recently discovered Kathryn Porter (linked above) and she talks a lot of sense. It helps having degrees in BOTH Physics and Economics, she is the sort of person who _should_ be energy minister.
Building datacentres at this pace is worrying, because they are a constant load day and night, wind or calm. And things like "demand control by voltage reduction" (one of the tools in NESO's toolbox to avoid future blackouts, so they say) doesn't work on them, because their electronic power supplies would simply draw more Amps if you gave them less Volts.
At the same time, we have crackpot NIMBYs objecting to "pylons", believing that everything can be underground or offshore. Unfortunately they miss a few key points: 1. Just because London has some short (and exceedingly expensive) 400kV AC cables in tunnels (see: London Power Tunnels, cost: £2bn), doesn't mean that it can be done over long distances. Even if we could afford the tunnels and the cables, the capacitance of a HV AC cable fundamentally limits its length. It becomes inefficient, and you need very expensive shunt reactors to correct the power factor, even for short distances. 2. HVDC is great for cross-channel links between two separate grids, but it is a terrible idea internally. It can't work without stable AC commutation at both ends, and doesn't contribute to grid stability. If the 50Hz goes out of tolerance by 1Hz, HVDC links along with most Wind and Solar will trip offline. 3. They say "Just uprate existing pylons" - that is being done, with High Temperature (HTLS) conductors that can withstand operation up to 250°C at double the normal load for their weight, but this is inefficient, expensive, and doesn't improve system reliability - anyone on these forums knows that two lines of pylons have more resilience than one.
Perhaps anyone wishing to build a datacentre should have to pay their share of the next nuke plant, and the transmission infrastructure to go with it
A heroic effort indeed. However the response received from minitru^W ofcom is hardly comforting..
> The Act makes no mention of how closely connected to the provider content the comment or review must be to benefit from this exemption
So.. This is an IT forum and we are (in the main) discussing IT related matters, with the occasional bootnote. The Act still leaves it in the hands of ofcom to decide if we have deviated from our remit, essentially giving ofcom the potential power and remit of "Roskomnadzor".
You mean the one to steal crypto keys from er, at least 7 people?
Although yeah, not really taking off, which is fortunate.
What I mean is, said crypto scheme ought to be investigated as a possible source of the fake package. If they are the sole project on GitHub using it, then it is more than a little suspicious that they might not be innocent victims.
I have never (and never intend to) work with Amazon S3. But what you are saying is, Amazon lets you address a storage "bucket" by name and name alone, and then when you stop paying your money, they will let anyone else register with the same name and start stealing data from your dead Android apps etc??
Shirley, they should not allow name re-use, not without some proof that you are the original owner
Or at the very least, there should be some mandatory API key authentication, and the keys obviously would be regenerated when the bucket of the expired name is created by someone else
There are two things which can't be reloaded without a reboot:
The static-linked part of the kernel i.e. you can update kernel modules without rebooting, but not the kernel itself. (Ok there is kexec, but i'd be surprised/interested if anyone uses that in production)
Then there is PID 0 (init) which can't be killed on a UNIX/Linux system
Both of these need to be made as simple as possible to avoid the need for reboots
And that's why I find it so perverse that PID 0 is picking up so much bloat in systemd
Can you link to the forum post that describes your issue?
TBH as mentioned earlier i have little personal experience with Mint and always use Debian. It has a very reliable install process for me, but still it could perhaps be offputting for the average member of the public, because it is designed to be flexible for power users
Then there must either be a hardware problem or it's a very unusual rodent. What 'driver' did you try to install? (It shouldn't need one, and trying to install a special one could be why it is broken).
It should show up in 'lsusb' as a USB Human Interface Device, and 'hd /dev/input/mice' should print hex to the terminal when you move it. If all that is as normal, then it's perhaps an issue with the display manager and not the mouse. If you are using Wayland, try xorg instead
In any case, i'm surprised and disappointed, and frustrated because whatever it is is fixable.
Do other mice work on the same system? Does this mouse work on other systems (without special drivers)? What mouse is it?
Edit: ignore all that, just read your other post, glad it is working
Bluetooth mouse? You probably want the Bluetooth menu rather than 'drivers'.. And you need to put the mouse into it's pairing mode, etc. It can be a faff initially, but no more so than on Windows, at least when Bluetooth itself is working. (My experience is with KDE on Debian, and it my Bluetooth MX Master 'just worked')
Any mice i have used have been either USB HID (no driver needed) or Bluetooth, what is yours?
They couldn't resist the opportunity for a few 'explosive' flatulence-based puns, but I agree they parped on about it a bit too long
Then again, the Reg often has articles about Wind and Biogas, this could come under their unclean energy topic
And in any case, it is in the erm, buttnotes section.
.. I'll get me coat
They always have been.. See Surface RT, Windows Phone, etc..
Useless, compared to an AMD or Intel device equipped with Linux/Wine, or indeed an Arm device with no trace of Microsoft shitware whatsoever, for those with no need for legacy software
They are trapped in a valley of doom with this, I fea.. rejoice
Frankly, I am a just a little bit worried that the attendees of Trump's inauguration (Billionaires who are already spending Billions on constructing their own personal bunkers in New Zealand, and perhaps Greenland next) might have a dastardly plan inspired by Hugo Drax of Moonraker fame, whereby they see the world as having too many humans in it, and wish to start World War Three (which they are certainly in a position to do) and retire to their bunkers with some 'perfect specimens of humanity' so that they may inherit the Earth
Said apocalypse could be survivable if killer robots, AI engineered disease, and/or sudden technological collapse were used, rather than nuclear weapons.
