One has to put this in perspective. On the rdos "Aspie Quiz" (which, at least of the previous version, followed the official diagnostic procedures for autism extremely closely and accurately measured "autism levels"), I score 178 out of 200, well into the upper range for autism. I've been officially dxed with autism and complex ADHD. Amongst a bunch of other stuff.

I hyperfixate (though generally not on Slashot, interestingly, although again there are exceptions), and my language will, at times, get blunt. And, yes, have been known to do all the other things you list. Although I do make some sort of effort to keep it at levels others can tolerate. Sometimes, I even actually succeed in this.

As a result, I think I can reasonably and fairly say that autistic people generally don't fit rsilvergun's profile. In fact, I suspect that the number of people on the ASD spectrum on Slashdot is well above the background level and quite plausibly much higher than in even the sciences. I could be wrong, there, of course. That does occasionally happen(!). So "autistic" (even "severely autistic") doesn't reflect actual behaviour in quite the way that the "standard image" portrays.

And that's one of the biggest alarm bells you can ever have with this condition. I've been in autistic groups where half the participants can't ever leave specialised care, and even those never ticked all of the boxes. If you see someone who DOES tick all the boxes, it is of course possible that they are autistic, but the underlying neurology of the condition (which is highly complex) strongly suggests that they can't have all those behaviours because of autism. Almost certainly, at least some behaviours are a fiction, even if it's not easy to figure out which ones are real and which ones aren't. And if some of them aren't genuine, you can't trust that any of them are.

Remote diagnosis is a dangerous game, but if someone exhibits two symtoms that appear in a description but cannot actually coexist, that's the time to stop trusting what they say.

If you want to stream and store every single episode of Thunderbirds in 1K, you're welcome to try. Although International Rescue might stop you.

(It's a pity that the 4K upgrades they did on two episodes weren't popular in the cinemas - the quality was impressive and actually showed just how much effort was put into making high quality models even for a cheap show in the 1960s. You couldn't upscale the early Doctor Who stories to 4K without a LOT of cleanup, the props weren't nearly to the same standard.)

Streaming is inherently quality-capped - there's only so much pipe coming out of the streaming service, it's gotta handle an Internet clogged with cats and porn (and, trust me, you don't want the cats in the Interwebs batbatbatting your film to knock it over the edge), and it's got to be a simple enough format that low-end low-power laptop/phone CPUs can handle it.

So it's partly watch-forever for DVDs, but also a case of what to do if you really really want high quality.

Agreed, format wars were stupid, but DVD has limited capacity and there's not much you can do about that. DVD-9 is fine, but heavy on the CPU, and (whilst backwards compatible) is nonetheless another standard.

I dunno. You might watch low-quality stuff - I dunno - but there's plenty of high-quality productions where bluray (even if it's not 4K) offer a definite advantage over streaming. Audio is also much higher quality streamed. Heavy compression may be ok if you're not used to anything decent or not watching anything decent, but high quality sound is always going to win for me.

Crowdstrike has such a strangehold on corporate IT that we will only see more - not fewer - occurrences of their software itself taking down networks. We all remember the fairly recent event where crowdstrike did billions of dollars in damage to networks around the world with a faulty update. Since then even more companies have adopted it.

I work with many large companies who run crowdstrike. When I ask their IT folks how it works - or even how it is configured - I get blank stares back. Presumably someone knows how to configure it, but that someone is never the person I get to interact with. If I'm doing an installation and just need to connect a USB drive to a new PC it can take hours just to get permission to do so. If I install our software first (before connecting the new PC to their network at all) and then they install crowdstrike, crowdstrike can render the PC completely unusable without warning - leaving us no choice but to nuke the PC and start over from the OS installation. If they install crowdstrike first it might lock out so many ports and services on the PC that I won't be able to install our hardware and software at all.

Again, virtually nobody on the IT staff know how to handle the issues. I'll spend hours at the keyboard with them, with them using various admin accounts, and we won't get anywhere. And there is no way to predict which setups will go sideways with crowdstrike installed first versus which will go sideways if it is installed later.

One important thing I have learned - crowdstrike updates and policies are far, far from instantaneous. IT will install them and it may be an hour or more for everything to take effect as the updates and policies come down from the server. Something that works at 2:30pm might suddenly be irreversibly broken at 2:40pm, without warning.

This is not how IT security should work.

The 2.(stable/devel).patchlevel format worked extremely well and stopped version number explosions. The main drawback to it was that it was prior to git, and so the patchlevel could get very high. We also don't need stable/devel, any more, as we've now got one tree for stable and a different tree for devel.

Having said that, I did very much like the three digit split, even though (as Linus as repeatedly said) it was something of a fiction at times. We do sort-of have that, now, with the third digit being used to mark backported stability fix rounds. And, yes, I would agree that version numbering is a fiction of sorts anyway.

I really don't like the major number incrementing at the speed it does, though. Yeah, 3.5 years between a major number increment is sort-of ok. That's 42 months, and 42 is indeed the answer to life, the universe, and everything. And an OS kernel isn't. in all fairness, really susceptible to being divided up into the major.minor.patch format because none of these really mean anything in this sort of a context.

Dunno how you'd really go about improving the system.

Linux removes a feature from a kernel under one of two conditions only:

1. The feature isn't maintained any more AND is now so stale it cannot compile AND nobody is willing to take on the work to make it work
2. The feature refers to hardware that is so obsolete that the number of users is effectively zero insofar as anyone is capable of determining

As a result, you're generally safe with anything that is built into the official Linux kernel tree. The API provided to applications is incredibly stable and Linus reputedly has an army of dedicated berserker Vikings enforcing this.

However, binary-only drivers and non-standard components are another matter, as they're maintained out-of-tree and don't always comply with Linux kernel practices. This is the only area you have to be careful, as distros aren't always clear as to what is official and what is stuff they've grabbed off the net and linked in.

Notice that the only value he sees in Humanity is as "useful and compliant factory workers", for times-adjusted values of 'factory'.

He likely considers a family reunion picnic to be a waste of resources.

"ChatGPT never needs to put its daughter into piano lessons - think of the savings!!!" - probably.

IIRC ZFS 2.4 will do parallel zpool imports, so that should help a bit, especially with a big server.

I think half my boot time on a big backup server is LSI device enumeration - which is totally not parallel - so there's only so much one can do.

Could the mptsas devs parallelize this task? It would certainly make lots of people happy.

It's also possible I could fiddle with systemd dependencies to get me a login shell before all the pools are imported because those pools aren't needed at all for the root filesystem. They'll be used minutes to hours later when a backup is initiated.

On my media machine some services like jellyfin need certain filesystems up before it starts but I don't need jellyfin to be started before sshd comes up. Etc.

We're still a ways from having a good tool to automate these dependency trees and most sysadmins are "meh, I can wait two minutes". It's certainly not the ideal for remarkably fast computers.

Also being starved for PCI lanes on pretty much every system is a pain. I'd take lanes over more cores or more GHz any day of the week. Let the data flow!

Why bother? All you need to do is to DNAT at an intermediate point and identify a nearby doughnut shop as a repair facility. The F35s will then land there and you can reprogram them whilst the pilots are stuffing themselves.

Oh. No. Wait. That's cops. Cops don't use F35s (yet).

Always always always remember that Loki can beat Odin.

It depends. If the cats of the world decide to rise up in revolution (which, given the quality of catfood, is entirely plausible), then being able to control all the vacuum cleaners would be the only defence.

> Why was it removed in the first place,

Per Rene:

IMHO The removal of XAA was a huge cooperate planned obsolesce mission for older GPUs. Rendering everything mostly unusable slow. even for period correct X11 apps. The code should have just been left in peace and only bugs and security patches applied instead of outright deleting it for no good reason.

Here's a commit for XAA support in T2 Linux for others who are interested. I hope Rene has time to push it up to XLibre since it seems like the Xorg people are going to steamroll Wayland if they can and the XLibre fork will be the only surviving X11 server. Obviously it would be best if every distro could run on older hardware and Wayland is likely a poor choice for vintage computing.

I didn't know about T2 Linux and it really looks fantastic - I thought NetBSD was my only choice on some of those machines. Some of the screenshots feature WindowMaker, the spiritual successor to NeXTStep, which ran on an '030 and 2D video so this all makes perfect sense.

Those machines were perfectly usable and we can actually afford, today, the amount of RAM they used.

The claims against the archive owner are wild and would be easily disproved if untrue.

Is this the same operator who would block readers if their ISP used some DNS feature he didn't like, back in the day?

I understand being disagreeable, but, jeeze, this takes it to a whole new level. Way to have people's sympathies and then burn it all to the ground with malice.

Wikipedia was apparently in the position of being forced to amplify the attacks with their links to the archive. Not a supporter of theirs these days but what else were they to do?