Denvercoder9

> Other platforms find better solutions.

Other platforms make different trade-offs. Most of the pain is because on Debian, it's customary for applications to use system copies of almost all libraries. On Windows, each application generally ships their own copies of the libraries they use. That prevents these incompatibility issues, at the cost of it being much harder to patch those libraries (and a little bit of dikspace).

There's nothing technical preventing you from taking the same approach as Windows on Debian: as you pointed out, the libc ABI didn't change, so if you ship your own libraries with your application, you're not impacted by this transition at all.

> They have broken the userspace ABI for lots of libraries again.

If the old ABI used a 32-bit time_t, breaking the ABI was inevitable. Changing the package name prevents problems by signaling the incompatibility proactively, instead of resulting in hard-to-debug crashes due to structure/parameter mismatches.

It's not.

I can't reproduce the exact datapoints from the site using `git grep`, but most of it seems to be down to a single commit that removed repeated usage of fuck from one file: https://github.com/torvalds/linux/commit/a44d924c81d43ddffc9...

It's actually because someone with "crap" as a substring of their e-mail address made a bunch of contributions with their e-mail address in it (e.g. in maintainer records and copyright notices) around that time. Nothing to do with COVID-19.

See the graph with the entire domain for comparison: https://www.vidarholen.net/contents/wordcount/#crapouillou

> they're making a final movie.

They're not making a movie; the entire series is a prequel to the 2016 movie Rogue One.

The C++ standard doesn't forbid introducing side channels, so the answer to the question is yes.

It depends entirely on the context. For routes where total travel time is mostly governed by moving time, and the stationary time in stops is negligible, the capacity boost from double-deckers easily outweighs the longer (un)loading times. The alternatives to increase capacity can also be problematic: with longer trains you start running out of platform length (and long platforms add walking time); while running more trains closer together requires more personnel and rolling stock, and is limited by signaling block size and braking distance.

> why would they have their system language set to X if they speak Y? If they want Y, they should just set their system language to Y!

If only they respected my system language. All my language settings are set to English, yet I routinely get autotranslated crap to my native language.

ProxMox supports it, but it's not what the linked script does, nor is it officially supported by HAOS.

> Proxmox adds very little overhead.

It's still running a second kernel and entire userspace stack. In my world that's not "very little overhead".

> What I wanna know is if I could have a plug-in box that detects the frequency is drifting

Yes, it's not that hard. There's smart meters and plugs that have frequency measurement built in.

You can even do it with an audio cable: https://halcy.de/blog/2025/02/09/measuring-power-network-fre...

> They could live with an older version of GCC for a year.

That's just not what Fedora is, though. Being on the bleeding edge is foundational to Fedora, even if it's sometimes inconvenient. If you want battle-tested and stable, don't run Fedora, but use Debian or something.

The article is terrible, this doesn't affect the default NAPI behaviour. See the LWN link posted elsewhere for a more detailed, technical discussion. From the patch set itself:

> If this [new] parameter is set to a non-zero value and a user application has enabled preferred busy poll on a busy poll context (via the EPIOCSPARAMS ioctl introduced in commit 18e2bf0edf4d ("eventpoll: Add epoll ioctl for epoll_params")), then application calls to epoll_wait for that context will cause device IRQs and softirq processing to be suspended as long as epoll_wait successfully retrieves data from the NAPI. Each time data is retrieved, the irq_suspend_timeout is deferred.

> If you have a home machine doing network routing it would absolutely benefit from this.

It most likely won't. This patch set only affects applications that enable epoll busy poll using the EPIOCSPARAMS ioctl. It's a very specialized option that's not commonly used by applications. Furthermore, network routing in Linux happens in the kernel, not in user space, so this patch set doesn't apply to it at all.

The study you're quoting also says that roughly half of the remaining 81% thinks that God has guided human evolution, so it does contradict OP's statement of 40% believing God created the Earth 10,000 years ago at all.

> I can’t sell my family member my house for $1 for example

In most countries you absolutely can. The difference between market value and the sale price might be considered and taxed as a gift, but such a deal is generally not prohibited.

Or not in the US.

> The really egregious thing is that when businesses pay for Slack, it remains unindexed and they just change the retention to 1yr.

That doesn't track with my experience as a user at all. Almost every day I do a search that returns results older than a year.

A better solution is to not tie visas to jobs. While at it, also stop tying healthcare insurance to jobs.

As someone from a country with strong labor protections, I'm in favor of making it easier to fire people. There's too much dead weight around in companies here. But being fired shouldn't be a life threatening event.

> A business (supplier) that doesn't sell to end consumers pays no VAT

They indeed pay no net VAT (it's not a cost for them in the sense of their profit and loss statement), but they do remit a bit of the VAT collected by the end consumer to their _local_ tax authority.

As an example, let's consider a VAT rate of 20%, and a Dutch company that buys from a French one and sells to a German one. Their costs per product are €80, and thus they pay €16 of VAT over that to their French suppliers. If they sell a product for €100 (i.e. they add €20 of value), then they collect €20 of VAT from their German buyers (which might in turn get it from the end consumers). There's a difference of €4 between what they received and paid in VAT, and that difference is collected by the Dutch tax authority. That €4 is not coincidentally the 20% VAT over the value added by the Dutch company.

It's not an engineering problem, it's a political problem. If you present this problem to any power engineer at Nvidia, they'd probably say something akin to "yeah, delivering 600W at 12V over a 12-prong connector is insane, up the voltage". The issue is that 12V has been the standard voltage for ages, and if you want to sell a product that requires a higher voltage, you first need to get the industry to produce PSUs that deliver a higher voltage.

> A 48V system would have only ~13% the resistive losses over the cables (more importantly, at the connections!)

It's one-sixteenth (6.25%) actually. You correctly note that resistive losses scale with the square of the current (and current goes with reciprocal voltage), so at 4 times the voltage, you have 1/4th the current and (1/4)^2 = 1/16th the resistive losses.

Because there is a non-zero chance that at some point you'll need a medical procedure that is, even at non-inflated cost price, more expensive than your net worth in order to survive; and society considers it unethical to let you die in that case. With the same reasoning you're not actually the single payer for your food, and food stamps exist.

In case you don't know, you can use the `-n` argument to `systemctl status` to tweak the log output, e.g. `-n0` to disable the log output and `-n40` to get more than the default 10 lines.

> I am annoyed by the redundant "systemd/system" directory name

It's not redundant, you also have /etc/systemd/user (and /lib/systemd/user) where units that run in the user context (as opposed to system-wide context) are stored.

Second stages definitely are getting dropped elsewhere, commonly the southern Indian Ocean, as well. Point Nemo doesn't always or even often line up with the target orbit, and you can't keep second stages in space for extended periods of time, because the propellant needed to deorbit boils off.

It's not space debris, it's the deliberate disposal of the upper stage of the rocket precisely to prevent it from becoming space debris. The time and location of re-entry are planned and controlled. This is not going to crash into your neighborhood (except if you're neighborhood is in certain areas of China, where they they happily dump spent rocket stages on populated areas).

> unix time is discontinuous around leap seconds.

Wrong. It's continuous, it's just not monotonically increasing.

Not OP, but with supercapacitors you're talking about a wildly different amount of energy. A 5F/5V cap stores about 62 J, while a 2500mAh 1.2V AA battery stores 10 kJ. With three of them you can store almost 500x as much energy.

> So about 2/3 of the latency is caused by relays and switching equipment.

Not really; the speed of light in fiber optic cable is only about two-thirds of that in a vacuum. That means it takes light about ~60ms to travel the 12,000km great arc distance.