johnklos

Running Linux on a 4004 is possible, as we've seen, but running llama is just way too far? Interesting take.

Yes, a Macintosh LC II :)

Dmitry talks about compiling the kernel in years. While I haven't done that, I have built NetBSD-vax and NetBSD-mac68k natively on a VAXstation 4000/60 and on a Mac Quadra 610. It's on the scale of multiple months instead of years, but it's enough to give me a feel for it.

Wonderful look in to our history, and you're undoubtedly correct about being able to target that system.

My example was about hardware that was affordable to mere mortals (although it's getting to be more expensive to buy that same hardware now as it was to buy it when it was new), but the idea is the same :)

I've always interpreted the definition of storage as arbitrarily large, not specifically infinite. The universe, after all, is finite. The "well, acshually" arguments aren't interesting, because they're 100% abstract.

There are many, many things that can't be hosted in Docker. For instance, literally everything that I run on many of my servers can't, because Docker only supports certain platforms and architectures. Things that we want to run on systems with modest amounts of CPU and memory can't be run via Docker for obvious reasons. Stuff that needs hardware access can't. Stuff that needs finer resolution of permissions can, but there's a lot more work that needs to be done.

But even ignoring those, if I'm going to spend all the time needed to containerize everything in to Docker images myself, why wouldn't I just run the programs directly and not deal with the overhead and extra work?

I installed Windows 95 on an Amiga 3000 with a 25 MHz m68030 via floppy to see if DMF formatted disks would work and to play around. By the time it finished, I had forgotten what I wanted to try out.

Wow. And I thought modern NetBSD on a 15 MHz m68030 with a 16 bit memory bus and 10 megabytes of RAM is slow. This is crazy!

It illustrates a point I've explained to many people over the years: once computers started coming with persistent storage, open address spaces and MMUs towards the late '80s and early '90s, we basically arrived at modern computing. An Amiga 3000 or i80486 computer can run the same things as a modern computer. Sure, we have ways to run things orders of magnitude faster, and sure, we now have things that didn't exist then (like GPUs that can run code), but there's no functional difference between those machines and new ones.

I love that Dmitry shows how loosely "functional" can be defined :)

It is a bit easier, at least. My almost 90 year old Mom now knows to be suspicious of email and to not believe email unless she has a reason to think she should be getting it.

To be fair about setting up a Pihole or some other form of DNS filtering, that's something that the network administrator should do, not individual users. It's a shame that it's still not trivial - companies that make NAT routers resist building in things that they don't completely control, so a configuration page for Pihole in your NAT router's web interface likely isn't coming soon. I hope that changes.

Mom also understands that someone taking over her Nextdoor account would be a nuisance, whereas someone taking over her banking account would be significantly more problematic, so the more important something is, the more time she'll take to ascertain its authenticity.

I practice explaining these things because I do it often. One interesting observation is that Mom believes me, so she does the things I suggest, whereas younger people think they know better, so they generally don't put much energy in to my suggestions. I'm working on ways of showing people that they're not necessarily safe because they're "doing the same things they've always done, and nothing bad has happened yet".

Cloudflare's abuse form will not let you submit the report if you don't include a URL that currently points to their network. There're no options for phishing / scam domains for which they're the registrar and/or DNS hosting.

Can be summarized with: Don't click on links in email.

So is github-scanner.com (and github-scanner.shop) still the same malicious party? It seems to be. Funny that their DNS is hosted by Cloudflare (who, famously, don't host anything, because they think we're all dumb). Cloudflare, who take responsibility for nothing, has no way to report this kind of abuse to them.

The domain which hosts the malware, 2x.si, both uses Cloudflare for DNS and is hosted by Cloudflare. At least it's possible to report this to Cloudflare, even though they rate limit humans and have CAPTCHAs on their abuse reporting forms.

Sigh. Thanks to Cloudflare, it's trivial these days to host phishing and malware.

You don't. Google doesn't have humans that non-paying customers can reach.

As I wrote, either you work on improving your reputation (set up all the acronyms, send lots of test messages, tell everyone you know that uses Google for mail to look in their filtered mail and tell Gmail that your mail is not junk, keep sending, et cetera), or you smarthost through a reputable service.

You're right that I didn't check them. I said that.

It's like doing research for colo, like my example. If you have the need, then a couple of hours of research is well worthwhile. I don't have the need, so I'm not going to do it now, but that's how one starts.

The colo example is apt - colo providers that don't have pricing are invariably too expensive, so I skip them, but there are plenty of others to check out that aren't Cloudflare. The one article I skimmed even says whether the providers are pricy or affordable.

Nobody needs Cloudflare. If (most) people were aware of how much Cloudflare breaks visibility across the world, they'd likely avoid Cloudflare, too.

That's definitely not true. You can either put in the work to improve your deliverability, or you can smarthost through a provider that has a good reputation.

"to self-host any app"

Docker != app. Perhaps it'd be more accurate to say, "to host any Docker container"?

I don't use them myself, but I only choose colocation providers that have a good handle on their own protections. A quick search, though, shows lots of reviews and options:

https://www.techradar.com/news/best-ddos-protection

https://www.gartner.com/reviews/market/ddos-mitigation-solut...

https://expertinsights.com/insights/top-distributed-denial-o...

No idea about the content of those links, but considering the amount of research I do before selecting a colo provider, it'd be trivial in comparison to research a DDoS protection service.

You should design your site to be resilient to spam traffic, not try to filter until it's gone. By filtering, you've become unreachable by much of the world, spammers or not.

You can self host email. There are plenty of how-tos, prepackaged packages, even books:

https://mwl.io/nonfiction/tools#ryoms

I'm writing my own how-to which includes finding suitable Internet providers and includes setting up smarthosting in case you can't get a good address for yourself.

Anyone who actually runs machines can tell you that the cloud is an expensive way to run anything. If you have zero servers, then sure, running something small might be cheaper, but if you have any number more than zero, running things on your own server(s) always make more sense.

It's interesting that the people who bought in to the marketing believe that you no longer need systems administrators - now you just need someone who intimately knows AWS and their billing shenanigans, and who will babysit used resources to avoid incurring huge bills.

Those marketing fans also say that the cloud is necessary for scaling up, as though they could never imagine running services on owned servers in addition to maintaining services in the cloud that can be turned on and turned up when needed. When people are that blinded by marketing fluff, you know that their cloud expenditure is going to be much higher than it needs to be.

I've been helping people move from the cloud to servers they already own for years. The cloud has its place, but anyone who thinks everything belongs there shouldn't be trusted with tech or with spending.

Microsoft has an app called, "Windows App". Perhaps this shouldn't be surprising from a company that named their OS after a single featur, but this name is so much worse. What does it do? The name doesn't say at all.

Did it really occur to nobody at Microsoft that "now open the Windows App app" is going to confuse a lot of people?

Even after reading the whole article, all I can infer is that this is a "seamless" replacement for Microsoft's "Remote Desktop". I'm assuming it's 100% Microsoft cloud based, although I'd be curious if it can be used without the Internet.

There are *so* many options out there. Saying you don't know how to do it without using an evil, monopolistic company is like saying you can't host email without using Google. It's lazy, untechnical and just plain untrue.

"since you must process IP addresses in order to serve a website"

That's complete nonsense.

You listed just one slightly onerous requirement: allowing people access and agency over their data. If you don't store their data, you don't have to do that.

It's a bit hyperbolic to say that you're, "not even half way through the article and I'm skipping over tons of what it's saying needs to be done", when you've literally only listed one thing.

...and have worse quality, which makes the idea of caring about using real-time transcoding to give the best quality pretty silly.

I see both sides. While "self hosting" has always meant hosting yourself, and hosting on other people's systems isn't hosting something yourself, I can see how people can get confused and can call running their self-configured software on a rented VM "self hosting".

It's not as unambiguously incorrect as other silly things people say and do that are technically incorrect, but it is annoying when people don't provide enough context where it matters.

Honestly, the distinction only really matters when discussing privacy. Hosting your own stuff in a rented VM is still self hosting, but if you're talking about how you self host because you care about the security of your data, you're now definitely not talking about rented VMs.

Generally, I think we need to get used to the idea that "self hosting" now also refers to hosting software you configure on rented systems / VMs.

It's a good writeup, but I do take exception to this:

"Seriously, else-hosting is the practical option, let someone else worry about the reliability, concurrency, redundancy and availability of your systems."

Spend one time trying to get through a maze of automated phone answering systems, then try to ascertain whether the human, when you finally get them, even understands the issue, then wonder how much of what they're telling you is to just get you off the phone, all the time wondering if calling even really does anything, and you'll wonder whether it's better to blindly trust a company that likely doesn't have tech people we're allowed to talk to or to just do it ourselves.

At least when there's an issue with my things, I can address it. Although a bit of a tangent, I'd love to see a review of major hosting providers based on whether you can talk to a human, and whether said human knows anything at all about Internet stuff.

Some of us don't download and run binaries. For some of us, the advantage of ARM is exactly that it isn't x86.

Transcoding using hardware acceleration is one of those things that some people insist upon, like wanting "gaming" everything and overclocking everything, but most people don't need.

First, the quality of output of hardware accelerated transcoding is in most instances noticeably worse than CPU transcoding, as in most people can discern the difference watching two videos side-by-side.

Second, CPU transcoding is more than enough to keep up with most streams. For instance, 1920x1080 h.264 streams can be done in real time by even a modest 4 core processor. People who want 4K h.265 can buy a fancier CPU.

Third, most people are served perfectly fine by pre-transcoded files. Your comment about subtitles isn't correct - most streaming devices can accept subtitle metadata, and even if subtitles need to be burned in, they can be burned in to a static file and/or generated in real time by a CPU.

What's interesting is that the same people who want Intel Quicksync claim they want to transcode to the best quality that a device can support, then say they don't want to transcode to a static file because they don't want to transcode to a lower resolution / bitrate than what their best device can support, but want to be able to view things at the highest quality wherever they want. They've been tricked, because if they really cared about quality, they'd care about the difference in quality between QuickSync and CPU encoding.

That's why I tell people that if the really want highest quality real time transcoding and have sources that are ultra high quality, they'd be much happier with a Ryzen 7900 than with Intel QuickSync. Side-by-side comparisons would bear this out every time.

True. I've wondered the same thing, and always come back to inexpensive Ryzens on ASRock motherboards that I simply underclock.

There are lots of people out there saying to get one of these low(er) power Intel systems instead of, say, an Orange Pi 5 or some other ARM machine, but I think some of the information from the comments tells us a little more about their suitability for certain applications:

"In fact ‘whole systems’ were benchmarked and not just Intel CPUs (under identical environmental conditions) since as we know in the meantime on Alder Lake-N Intel chose to throttle down the memory controller above a certain thermal treshold (sic) which affects all those benchmarks needing high RAM bandwidth or low RAM latency."

I thought the idea of having a CPU that runs at crazy speeds for the first minute or two, only to thermally throttle to something more realistic, was something that only applied to the marketed-to-gamers, high end CPUs, but apparently not. Low power CPUs should be simple to cool and should only ever thermally throttle in extreme cases (like running in a hot attic or car or something similar).

When I consider that an Intel N100 takes twice the power yet doesn't deliver even close to twice the performance of an Orange Pi 5, I can't help but lose interest.

Even comparing the N100 with my long time favorite low power x86, the AMD Athlon 5350 from ten years ago, I can't help but feel like Intel really hasn't innovated much here. I'd love to do some benchmarks, but I think the N100 isn't even twice the IPC of those decade old CPUs, yet those ran four cores at 2 GHz using so little power that their tiny little fans don't even turn on much of the time, even when all the cores at 100%. Where's your version of that CPU, Intel?