Posts by Crypto Monad

Airbus takes its long, thin, plane on a ten-day test campaign

"a target of approximately 100 hours flying time over ten days with no systems power-down"

They won't catch the Boeing bug where the plane needs to be rebooted after 51 days then...

Linux 6.6's in-kernel SMB networking server graduates

Re: Age Old Architectural Mistake Coming Home to Roost?

> getting rid of the networking to userland now, before it's too late

It's been done - see VPP and DPDK.

But I don't know if anyone has plugged in any endpoint network apps, as opposed to just routing packets.

Scientists trace tiny moonquakes to Apollo 17 lander – left over from 1972

I bet the original "250 degrees Fahrenheit" figure is +/- 10 degrees at least, even if that was at a single point on the surface.

Hence the translation to "121.1°C" is completely spurious, by about 2 significant figures. The "~" sign implies that this was rounded from an even more accurate figure!

From browser brat to backend boss: Will WASM win the web wars?

Re: Welcome back Java promise!!!

Sadly, many network and server admin tools were (or even still are) provided as Java applets. Examples include various server ILOs with Java-based remote console, and the Cisco ASDM management interface for their ASA firewalls.

The server vendors have generally moved to HTML5, but Cisco remains a relic of the past.

It was always a pain trying to get a compatible version of JVM and Webstart which worked with them, *and* which didn't violate Oracle's licensing terms (*), *and* which didn't reject the applets due to code signing issues or other problems.

Apple stopped shipping Java way back, from macOS 10.7.

(*) In many cases you would be prompted to download a newer version of Java SE. However, only Java SE 8 8u202 and earlier were free; anything later than that exposes your entire company to a per-seat licensing cost.

A license to trust: Can you rely on 'open source' companies?

I think Elasticsearch will be the one to watch here.

Does Elasticorp have enough loyal existing customers to carry them forward? The limited open source license arguably gives them a warm fuzzy feeling that they're actually getting something in return for their subscription.

Or is the Opensearch fork going to be where the real work is done, and ultimately tempt the paying customers away from Elasticsearch entirely?

Cost of gallium goes up after Chinese export restrictions land

Re: Retaliation

In the very arena you would expect responsible, level-headed people to operate, it's all being handled by a bunch of kindergardeners who only know how to throw the toys out of the pram.

True: but better a trade war than a real war.

Not call: Open source gurus urge you to dump Zoom

Re: how long before Jitsi disappears ?

Making your own git server be discoverable isn't a major problem. There are project home pages and search engines.

The problem is that Microsoft will discover it, and will still scrape the content to feed its AI.

Curiosity finds evidence of wet and dry seasons on ancient Mars

Re: Life's history

"They find some isolated spot with very few people around, then land right by some poor unsuspecting soul whom no one’s ever going to believe and then strut up and down in front of him wearing silly antennas on their head and making beep beep noises.”

Caltech claims to have beamed energy to Earth from satellite

Re: Misdirection

> 1 milliwatt per square centimeter

Converting to more conventional units, that's 10 watts per square meter. (Compare to ~1kW per square meter for direct solar irradiation)

This means you'll need a receiving antenna 100 times larger than the equivalent solar array working at peak capacity; or 10 times larger than one working at 10% capacity (e.g. on a cloudy day). There's a possibility it could work at night too, although the satellites would have to be able to keep out of earth shadow, and/or relay power between themselves.

At that power density, the "2GW" proposed from the article would need 20,000 hectares (200 km²) of receiver. Even if that can be done with a wire mesh that doesn't obstruct the sky, so the ground underneath can be used for other purposes, that's an absolutely vast construction.

And of course, you still need to buy the kilometer-square solar array, plus microwave power conversion. And more importantly, launch the whole mass into space. And then keep it in orbit (or replace it when it de-orbits).

AWS is running a 96-core, 192-thread, custom Xeon server

"We've asked AWS and Intel to confirm that spec is correct" - or just fire up an instance to see!

Also dmidecode (although less useful if it's not bare metal)

AWS: IPv4 addresses cost too much, so you’re going to pay

Re: IPv6-mostly?

Whether it's NAT44 or NAT64, it still requires CGN to reach the majority of the Internet.

NAT44 means they have to give your device both an IPv4 and an IPv6 address. With NAT64, they only have to give your device an IPv6 address.

By the way, you don't need DNS64 any more, if your device has an embedded CLAT - which is true for iOS and Android, and macOS if you send it the appropriate signals via DHCP and RA options. In fact, you can even open a page by IP address like https://1.1.1.1/ and it works fine over a pure IPv6 connection. The CLAT embeds the IPv4 address inside an IPv6 prefix used by your NAT64 device.

IPv6-mostly?

There's a nice write-up from RIPE about what happens if you run an "IPv6-mostly" network.

TL;DR: with modern Android/iOS/macOS devices it works quite well. If this level of interoperability had been specified 25 years ago, maybe we'd all be using IPv6 instead of RFC1918 behind our routers.

Unfortunately, since the vast majority of Internet content is still only on IPv4, you still need a NAT64 in your network. That's the fundamental problem: it means everyone still *needs* an IPv4 address, whether or not they use IPv6. But once you have an IPv4 address, there's little incentive to deploy IPv6, since everything of significance is already reachable via IPv4 (bar a few cat-feeders and Loops Of Zen)

NASA mistakenly severs communication to Voyager 2

Re: Off topic

> 300 baud is way worse than tedious.....I gave up on Prestel in short order!

When I used Prestel, it was 1200/75 (down/up). A system where I could type faster than the modem could send my keystrokes :-)

Intel adds fresh x86 and vector instructions for future chips

I was a big fan of 6800 and 6502, but IMO the pinnacle was the 6809. Two 16-bit index registers: luxury!

> Surely having 10 byte sequences just to encode new register references cant be good for performance and memory usage ?

It's an arguable point, but these days by far the slowest part is fetching data from DRAM, and that's most efficiently done by long burst reads into cache. Once the cache is warm, generally that's where the code will be executing from.

Some instructions being "long" are offset by common instructions being "short", and by the fact that you'd need to use multiple "short" instructions to achieve the same result as one "long" instruction.

> Isnt it time for a new cpu isa which stops the need for this terribly inefficient encodings ?

Well, ARM, MIPS and RISC-V are there for you to use. ARM seems to be doing pretty well at the moment: MacOS, Linux, Windows all run natively on it. But if ARM tighten the licensing screws too much, RISC-V could take off in its place.

Intel tried with Itanium, but failed.

Re: Far too many x86 instructions

With the fire extinguisher to handle speculative execution

Why do cloud titans keep building datacenters in America's hottest city?

Re: ... a capacity of 32 million megawatt hours....

Well, logically they could be talking about the total amount of energy the plant will generate during its operational lifetime. But I doubt it.

As a complete guess, they might mean that it generates 32 million megawatt hours *per year* - which would be about 3650 megawatts, on average.

What they say, and what they mean, don't necessarily coincide. But really, it's the job of journalists to get this stuff right.

UK government faces calls to end IR35 double tax anomaly

Re: IR35 is only "necessary" because dividends are taxed at a lower rate than earned income.

Perhaps the answer, then, is to tax Dividends (and Capital Gains) at the same rate as earned Income?

And perhaps add National Insurance on top as well.

I think it's a good idea - indeed to combine income tax and national insurance too. Then make earned and unearned income treated the same. I see two main sticking points:

1. The hidden tax of employer's NI contributions, where a company pays 13.8% of the salary for the privilege of hiring an employee.

2. Double taxation of corporate profits (i.e. company pays corporation tax, then the recipient of dividends also pays income tax)

make tax understandable

Hmm: that would be from the party which *didn't* introduce a 62% marginal tax rate for people earning between £100K and £125K? No no, all they did was taper the personal allowance for people earning over £100K.

Because that's completely different, right?

(The 62% includes 2% national insurance, now no longer capped, and is basically just income tax under a different name)

Goodbye Azure AD, Entra the drag on your time and money

Re: Meaningless Product Rename

> By ditching the AD moniker for ID they are able to ditch having to replicate in the cloud all the features they currently support on prem

I don't think they ever did replicate them in Azure AD. They had a completely separate product called "Azure AD Domain Services", where they basically spun up some VMs of Windows Server to act as your domain controller.

Ex-Twitter employees owed half a billion in severance, says lawsuit

Re: Nearly 600% reduction in staff and...

The baseline of the percentage is what you started with, not what you ended up with.

* If you start with a pound, put it on a horse, and end up with nothing, then you have made a 100% loss.

* If you had ended up with 50p, then you would have made a 50% loss.

* If you had ended up with £2, then you would have made a 100% profit.

Hence if you start with 1000 people and end up with 200 people, then you can either say that your staffing level is now 20% of what it was before, or that you have lost 80% of your staff.

Broadcom asserts VMware's strategy isn't working and it basically needs rescuing

Re: "VMware alone can't get that job done"

If it's that much of a [REDACTED] pile of shite, surely the regulator would be doing Broadcom a favour by not letting them waste their money on it?

Microsoft whips up unrest after revealing Azure AD name change

Re: To Marketing: a Boot to the Head!

Perhaps Microsoft could try marketing advice from plain-speaking Yorkshireman, George Whitebread:

https://www.youtube.com/watch?v=jY4tD2Hbg_A

Methane-spotting satellite that gives true readings of industry emissions hits skies in 2024

Re: The usual misinformation and obfuscation

> Surely we should be diverting all the funding from climate 'science' into the 'renewables' lobby so they can waste it building more windmills and solar?

In what sense is building wind turbines and solar "wasted"?

The reserves of oil and gas under the planet are finite: sooner or later we'll run out. We'd better have something ready in time for that.

And no, I don't expect nuclear fusion ever to give power which is cheap *or* clean, let alone both.

Re: The usual misinformation and obfuscation

I think the poster was trying to say that the CO₂ concentrations are high over the Amazon and Africa (*) - not the satellites.

And India is famous for its sacred cows.

(*) How true that is, I don't know.

InfluxData apologizes for deleting cloud regions without performing 'scream test'

Re: It seems like a mistake they won't want to repeat

That claim would have to be on some sort of *implied* guarantees of service - which might work in a consumer court, but not in B2B, where parties are expected to read the contracts they sign.

I haven't seen the contract itself, but I'd expect that it has clauses that explicitly disclaim all consequential damages; that compensation is limited to service credits; that they have the right to cease the service by providing X days notice; and that serving the notice via E-mail is treated as sufficient.

That doesn't mean that they were not jerks. At minimum they should have had brownouts prior to closure, identified the customers still using the service and helped them to migrate, and even at the end they should have *switched off* the service but not *deleted* anything for at least a month.

However, you can be sure that there will be no redress in court for those affected. Maybe some private deals will be reached, but only to sweeten those customers to stay with influxdb and not walk to a competitor.

Canonical takes its LXD 'containervisor' back into the house

Here's the backstory: the lead developer of LXD has decided to quit Canonical.

https://stgraber.org/2023/07/10/time-to-move-on/

Browser-based GUI

... is actually built-in since lxd 5.14. Nothing to install, just turn it on:

snap set lxd ui.enable=true && snap restart --reload lxd

I don't think we have to worry too much about it going in-house. Debian Bookworm has imported lxd already, and Canonical always were the main developers, and are claiming it will remain under Apache 2 licence.

NASA 'quiet' supersonic jet is nearly ready for flight

Re: As much political as technical

I used to live in Fulham in the late 1990's, and twice a day you had to pause for a minute as Concorde went overhead - it was impossible even to speak to the person standing next to you.

But we're not talking supersonic booms here - just regular engine noise. Those were *damned* noisy engines.

RAM-ramming Rowhammer is back – to uniquely fingerprint devices

Re: Why?

Plus: you'd have to allow the fingerprinting software to be installed on the computer you're probing (e.g. the thing which you suspect of being a bot). But if it's a malicious actor, it will fake its own fingerprint results.

I can only see this sort of fingerprinting being used by software vendors to enforce licensing - much in the way that they used to write intentionally bad sectors to floppy disks in the 1980's.

Ariane 5 to take final flight, leaving Europe without its own heavy-lift rocket

Re: All of our eggs are going into one basket.

However it would be thrown into the garden, and when you open the box you'd find the Atlas V has been replaced by a lorryload of cat food.

It's 2023 and memory overwrite bugs are not just a thing, they're still number one

"Clean and neutralize your inputs, people. You can't assume all your users are nice."

No: all user-supplied data can be handled safely. The real answer is, don't blindly interpolate user-supplied text inside some other specially structured text (e.g. SQL or HTML) without the necessary escaping.

Wrong:

sql.execute("select * from users where name='" + name +"';")

Right:

sql.execute("select * from users where name=?;", name)

Sanitising inputs, e.g. to reject single quotes inside names, is not the solution, and rejects valid values ("O'Hare" for example). You just have to handle your user data properly.

There are some interesting language developments to help enforce this, distinguishing programmer-supplied literal strings from user-supplied data. e.g.

https://docs.python.org/3/whatsnew/3.11.html#whatsnew311-pep675

Rocky Linux details the loopholes that will help its RHEL rebuild live on

Re: making them Red Hat customers, at least briefly

I expect the next step will be that the SRPMS will become extremely hard to get hold of: not included in the cloud base images, not available to 'developer' accounts, and only available to "Super Platinum Plus" customers on $10K+ subscriptions. (Essentially, under the same sort of terms that you can get Windows source code).

As for clouds, they don't even need to make the kernel binaries visible within the root filesystem. Hypervisors can boot from an external kernel image and initrd.

Chinese balloon that US shot down was 'crammed' with American hardware

"crammed" with off-the-shelf components that could have easily been purchased online

... and which were all manufactured in China, for American companies.

Virgin Galactic finally gets its first paying customers to edge of space

Re: Cheaper?

Virgin is far more expensive (the Vomit Comet costs about the same as a first-class transatlantic flight), but the zero G lasts a little bit longer.

I'm unsure about exactly which research projects are enabled by zero G that is longer than 20 seconds, but less than 6 minutes.

Bosses face losing 'key' workers after forcing a return to office

Re: employees were happy (31%), motivated (30%) and excited (27%) to be in the office

It would be clearer and more accurate to say:

* 69% of employees were not happy to be in the office

* 70% of employees were not motivated to be in the office

* 73% of employees were not excited to be in the office

Lawyers who cited fake cases hallucinated by ChatGPT must pay

Re: $5,000 Fine ... a Pittance for an Attorney

I would have thought that Contempt of Court would be a minimum offence here. I guess it depends on whether they were knowingly submitting made-up citations to bolster their case, or simply lazy and incompetent in not checking what ChatGPT returned. But even lazy and incompetent lawyers need to be punished.

The key point here is: ChatGPT is not a search engine. It's an engine for generating plausible-sounding text.

This is a real example:

"Yes, light is much faster than mobile signals. Light travels at a speed of approximately 299,792,458 meters per second in a vacuum, while mobile signals travel at a much slower speed through the air, usually in the range of a few hundred meters per second to a few kilometers per second, depending on the type of signal and the frequency used.

For example, the speed of electromagnetic waves used in mobile communication networks, such as those used for cell phones and Wi-Fi, is typically in the range of 3 x 10^8 to 3 x 10^9 meters per second. This is significantly slower than the speed of light and is due to various factors, such as the frequency of the signal and the properties of the medium it travels through.

While the speed of mobile signals is fast enough for most everyday uses, it is still much slower than the speed of light. This is why information sent via light, such as through fiber optic cables, can be transmitted much faster than information sent wirelessly through mobile networks."

This authoritative-sounding style is almost intentionally designed to deceive.

Red Hat strikes a crushing blow against RHEL downstreams

> These are EXPENSIVE products and they may never support non-RHEL installations

Somebody running an "EXPENSIVE" product probably won't object to paying a small amount extra to IBM for an RHEL licence.

But if the upshot is to start pushing the mainstream users into other distros, and then the vendors start supporting some of those distros too, it could be an own-goal for RedHat, with a smaller and less technically-savvy customer base. We'll see.

Gen Z and Millennials don't know what their colleagues are talking about half the time

“Let's think the unthinkable, let's do the undoable. Let us prepare to grapple with the ineffable itself, and see if we may not eff it after all.”

― Douglas Adams, Dirk Gently's Holistic Detective Agency

"You want to 'take it offline'?"

Or as used by the previous generation:

"You want to 'take it outside'?"

Can noise-cancelling buds beat headphones? We spent 20 hours flying to find out

Re: Old school here

Are the headphone an unofficial "do not disturb" signal?

No.

I will sometimes put a card in my shirt pocket saying "NO MEAL PLEASE", when I'm boarding a flight at 1am and don't want to be shaken awake by someone who thinks I want dinner at 2.30am.

Debian 12 'Bookworm' is the excitement-free Linux you've been waiting for

Re: About time!

A big USP of Debian 12 is that it will include a native deb-packaged version of lxd, a container management tool which gives you in effect "lightweight VMs"

lxd is developed by Canonical, and although originally it was distributed as regular .deb, they switched over to snap (which is also Canonical's baby) and dropped the .deb packages entirely, making it incredibly hard to troubleshoot. Debian now gives a way to have snap-free systems with lxd.

Ubuntu still leads the pack for ease of running ZFS on Linux though. You can do it with Debian, but your system will have to re-compile the DKMS modules every time you update your kernel.

The bonkers water-cooled shoe PC, hexagonal pink workstations, and IKEA-style cases of Computex 2023

IKEA cases

If you are looking for IKEA server mounting, the "Lack Rack" has been around for a long time:

https://wiki.eth0.nl/index.php/LackRack

Raspberry Pi production rate rising to a million a month

That was 800,000 in a *quarter*. That's only 3.2 million per year.

Windows XP's adventures in the afterlife shows copyright's copywrongs

Re: Coppyright length

Author said:

"Copyright terms are designed to give adequate payback to creators. Thereafter, the social value of free use was deemed to be greater than any residual benefits to whatever entity held the rights – that's why copyright has limits at all. This logic makes sense with creations that have a long commercial life – books, pictures, music, movies – but much less so for software."

I disagree with all of that.

Firstly, the idea that copyright has limits. Originally it did, but because of corporate greed and political arm-twisting, it's now as near as dammit unlimited.

Copyright terms are primarily designed to give payback to large corporations, and to make Paul McCartney incredibly rich. It's why Bruce Springsteen was able to sell his back catalogue for a reported $550m: investors buy it like an annuity which will generate a guaranteed income for many years to come.

However, I don't see that serves any useful social purpose.

Where is that income stream coming from? Obviously from the pockets of music consumers (e.g. subscribers to streaming services), but less obviously, by diverting money away from current, less well-known artists, into a small pool of incredibly wealthy artists and corporations. Money that *could* be paying this century's artists is going mainly to last century's artists.

Just like software, music and culture also has an incredibly short shelf life. Stuff created today is old hat next year. So why don't we let musicians and composers and other artists monetise their work for (say) 15 years, while it has some semblance of being "current" - and then make it free to use? Then the money pot can be focussed on those people creating *new* things.

And as for Paul McCartney: if he can't live on his billion-dollar fortune, he can either do some more concerts, or get a proper job.

Bookings open for first all-electric flights around Scandinavia … in 2028

> There's no reason to not believe that in 10 years, that 200km flight could easily be a 400-500km flight.

Except that the 200km plane doesn't exist, and is already predicated on assumed improvements in battery technology that haven't happened yet.

It's certainly true that recent *cost* improvements have been major: for a home solar PV system, you can now get double the kWh capacity for half the price, compared to even only 3 or 4 years ago. That may well continue.

However, a home 3.5kWh battery still weighs 32kg. Price is not the barrier to battery-powered aviation; weight is. Improvements to energy density have been comparatively minor.

> Predicating this entire process of transformation on a hypothetical future advance, when all of the currently hyped or anticipated "next big things" are adding a meagre few percentage points of improvement at best, is idiocy.

Agreed. Aviation counts for 2% of global emissions (or 8% of UK emissions), a relative drop in the ocean.

I wouldn't even bother with "offsetting" the last 10%: if we were emitting CO2 at 10% of 1970 levels, the planet will cope happily with that. Besides, most "offset" plans are a joke anyway - e.g. planting trees when they would have regrown naturally if you'd just left the land alone; or buying cookstoves to replace existing wood-burning stoves (which were carbon neutral in the first place).

However, aviation needs to show that it's sharing the pain, which is reasonable considering the breaks they currently get. No tax at all on aviation fuel?! That should be the first thing to change.

> The entire process of "decarbonisation" is economic suicide.

That I disagree with, and is why you didn't get an upvote.

In the last 50 years we have had massive economic windfalls; investing those back into solar, wind, storage and/or nuclear is just good sense, for climate stability and the long-term economic benefits that brings, not to mention energy security as the last year or two have made plain.

In any case, the supplies of hydrocarbons underground are not infinite, and could be put to much better use. I'm sure our descendants in a few hundred years will say: "we can't believe that you had that vast reserve of valuable chemical resources, and you just *burnt* it all!!"

Meta threatens to pull all news from California rather than pay El Reg a penny

Presumably those sites already have the ability to enforce their copyright, i.e. they can simply say "don't copy/publish our articles?"

I remember there was some legal disagreement around whether headlines were protected by copyright, but I don't think there is any ambiguity around ownership of article contents.

IR35 costs UK Research and Innovation £36M – the same it spent funding tech projects

Re: description chosen to make a point

I've previously contracted PMs that do this job a number of days per month ... Most of these people will only be doing this part of the time.

That may be true, but full-time or part-time doesn't really make any difference as to whether it counts as employment or not. Part-time employment is a thing.

It's down to other things like control, substitution, and mutuality of obligation. Roughly speaking: if you can instruct the project managers what to do or how to do it, then it's likely to be employment. If you contract them to manage a project, but they can do it using methods of their own choosing, or even subcontract the task, then it's likely not employment.

Microsoft and Helion's fusion deal has an alternative energy

Re: It's a sure thing

This.

If they had a fully working prototype today, which was stable and creating more power out than it takes in, then *maybe* in 5 years they could have the first commercial product (although that's not very likely) and even then it would be stupendously expensive to run.

But they don't. They have computer simulations, a failed prototype - the seventh one I believe - and lots of hot air, in various senses of the term.