> At some point you have to trust vendors. I don't know exactly how AWS guarantees eleven nines of durability on S3. But I sure hope that they do.
Trust is earned, it's built on reputations at the individual, corporate, and industry-wide levels. AWS has 20 years of reputation on which I can judge the value of their promises.
Not only has the LLM industry (it is not "AI" and never will be) absolutely not earned anything like that level of trust, the thing the technology has proven most effective at is in fact scamming. Making up something that looks/sounds convincing, especially if you aren't thinking too hard about it, is what they're best at. Combine that with a lot of money flying around and trust levels should be somewhere around "Elon Musk promises".
At this point there have been so many blatant examples of why you should never give a LLM "agent" control over production systems, but the allure of just giving some vague direction to a chatbot and telling it not to screw things up it just irresistible to some like Sideshow Bob stepping on rakes [1].
If everyone around you is whacking themselves in the face with the rake, and you know you can avoid it just by using your brain and not stepping on the rake, and avoid entirely by just keeping your rakes contained, but a rake vendor comes to you saying that instead they have built a new rake that they swear won't whack you in the face even if you leave it right in your walking path, do you trust them?
> No, IPv6 as it is supposed to be implemented gives (say) a single server a /64, which is for all intents and purposes an inexhaustible supply of IPs. You could in principle have an IP per site you visit and have plenty left to spare.
No, as it's supposed to be implemented a single internet-routable /64 is used per *network* and then most devices are expected to assign themselves a single address within that network using SLAAC.
ISPs are then expected to provide each connected *site* with at least a /56 and in some cases a /48 so the site's admins can then split that apart in to /64s for whatever networks they may have running at the site. That said, I'm on AT&T fiber and I am allocated a /60 instead, which IMO is still plenty for a home internet connection because even the most insane homelab setups are rarely going to need more than 16 subnets.
> So if I wanted to annoy GitHub, I could connect to them without ever using the same IP twice. Their response would have to be banning my /64, or possibly /56.
Well yeah, but it's not like it's exactly rocket science to implement any sorts of IP rate limiting or blocking at the subnet level instead of individual IP. For those purposes you can basically assume that a v6 /64 is equivalent to a v4 /32. A /56 is more or less comparable to /25 through /29 block assignments from a normal ISP, and a /48 is comparable to a /24 as the smallest network that can be advertised in the global routing tables.
> I rent physical servers from them that have been previously rented to other customers. At some point hard drives fail.
The comparison in this case is to Hetzner's VPS offerings, which are probably less powerful than the average "old laptop" but have a significant advantage in terms of hardware reliability. It's still possible for the host running the VPS to have problems which result in a crash or the VM equivalent of a hard power off but the VM hosts and their underlying storage should be redundant such that the virtual hardware never fails.
That's not to say rebooting from a crash-consistent state will always work, you should always keep backups even with a high-quality VPS host, but the odds of recovering cleanly from a hardware problem are orders of magnitude better than an old laptop. For the sort of hobby project or personal tinker box that would be reasonable to host on a random laptop shoved in a rack you probably wouldn't even notice the downtime until you saw the event notification email your provider sends you.
I've run a 7 figure business from an SSD shoved in a sata2 DVD-ROM slot in a DC because the end customer was being obtuse about upgrading from their "high end, best practice" raid 10 discs.
You use so many big words for nothing. All you need are backups. When it dies you restore. Nobody will care.
No, the Windows firewall in its default configuration does not restrict outbound connections in any way. Any application can make any outbound connection it wants. If an application attempts to listen for incoming connections from external sources and there is not an existing policy, Windows will pop up a dialog asking the user if they want to allow this and if so whether it should be allowed to listen on all networks, only networks marked as "private", or for domain-bound corporate computers only networks where the domain controller is reachable.
It can be manually configured with very detailed policies, but you have to know where to go to find those controls.
It's been a while since I used ZoneAlarm or Little Snitch, but the last time I used either one the default behavior was instead that any connection attempt or attempt to listen for which there was not a policy would result in a dialog showing all the details about what application is looking to connect to or receive connections from what as well as a variety of options for creating a policy or even not creating a policy and just deciding whether that one connection would be allowed.
Also back when I used ZoneAlarm I had dialup so the taskbar addon they had which showed realtime bandwidth usage and what applications had active connections was really useful. It also had a big red "Stop" button that would immediately disable all connections, which thinking about it in retrospect really makes me miss the more innocent days of the internet.
Not exactly the same situation, but RustDesk has recently been removed from the official WinGet community repository because their automated scans have been blocking updates since v1.4.2 in September 2025.
tl;dr: ESET Antivirus flags RustDesk as a "Potentially Unsafe Application" because it is a remote administration tool, despite not flagging similar commercial products in the same way, and the WinGet Community repo policy is to block anything flagged as such. Since they were unable to update the repo the RustDesk team requested that the older versions be removed to prevent users from unknowingly installing old versions that could potentially be a security issue in the future. Apparently this has been an issue for a lot of applications especially in the VPN and remote control categories.
There is a discussion about how best to handle these sorts of situations where legitimate and desirable applications get flagged as "potentially unsafe" or "potentially unwanted" but so far it's just been a discussion with no actual changes proposed yet.
4) Don't use a stack of plugins, if you must use any keep them as dumb as possible and stick to those with a longstanding reputation.
A basic instance, set to auto-update, installed on a shared webhost where OS/web server updates are someone else's problem is pretty foolproof. A VPS running a long-term distro set to auto update is almost as good.
---
That said I personally dropped Wordpress for static site generation years ago because I realized I didn't actually need any of the dynamic features and wasn't using the WYSIWYG editor. Now I write Markdown in to a file in a git repo and then trigger a regeneration whenever I update it.
This appears to be a fairly low effort simulation of a WeatherStar 4000, but it's been done much better by someone else here: https://weatherstar.netbymatt.com/
Also for those of us like me who grew up with a cable provider that still ran the old WeatherStar III, there's this one: https://weatherstar3000.netbymatt.com/
This may be an age or geography thing, but Weather.com/Retro looks a lot more like the weather channels I remember than either of your two links.
I don't think they were trying to recreate the thing you were expecting, but it seems a little harsh to call it low effort. It's neat! A fun little nod to their history.
> This may be an age or geography thing, but Weather.com/Retro looks a lot more like the weather channels I remember than either of your two links.
That'll be a memory thing then, because unsurprisingly there are some serious Weather Channel nerds out there and the history of the local programming as well as the STARs that powered it is quite well documented.
The only thing that looked anything like this was the WS4000. WeatherStar III and Jr. didn't have graphics at all while WeatherStar XL had a much more modern graphic design.
As others have stated, no, this isn't a simulation of WeatherStar. It does, however, look exactly like the weather channel did when I was a kid. (New Jersey and Maryland, in the 80s and 90s)
Because, IIRC (it has been a while since I’ve looked at the code) it grabs weather from the US National Weather Service… which is also a primary source of data for the original WeatherStar.
It’s useful for this purpose because all of the data is in the same format as the original down to the different forecast types and phrasing used
This reminds me of a sports channel I recall from the late 80s on cable TV. Sportsnet? Sportswire? I can't find anything about it online but I remember the music well.
Exactly. In the field of accident analysis and prevention the "swiss cheese model"[1] is very popular for this reason. It acknowledges that every layer of the system will have holes, and that ideally every layer when stacked up provides complete coverage. If something bad happens that means that holes every layer had to have aligned and the whole stack bears some responsibility for the outcome.
> the crazy ports of naughty things to naughty hardware (Netscape Navigator on Nintendo 64, oh my, how naughty you were, SGI!)
I remember in my youth when I first discovered Linux, soon after discovering that it ran on all sorts of architectures and starting to wonder how many of the computing devices I owned I could get running Linux.
The N64 and a Mac LC III were the only two I never managed to make it happen on. The idea of IRIX on one somehow never even crossed my mind, even though in hindsight it seems so obvious.
> Probably the best thing is a CB radio. Let them talk to any other kids in town but no chance of weirdness.
No chance of weirdness? On CB? Have you used a CB?!?
I had a CB in my car for a while and the majority of the talk I ever heard on it outside of traffic updates and cop reports on major interstate highways was weird shit.
Trust is earned, it's built on reputations at the individual, corporate, and industry-wide levels. AWS has 20 years of reputation on which I can judge the value of their promises.
Not only has the LLM industry (it is not "AI" and never will be) absolutely not earned anything like that level of trust, the thing the technology has proven most effective at is in fact scamming. Making up something that looks/sounds convincing, especially if you aren't thinking too hard about it, is what they're best at. Combine that with a lot of money flying around and trust levels should be somewhere around "Elon Musk promises".
At this point there have been so many blatant examples of why you should never give a LLM "agent" control over production systems, but the allure of just giving some vague direction to a chatbot and telling it not to screw things up it just irresistible to some like Sideshow Bob stepping on rakes [1].
If everyone around you is whacking themselves in the face with the rake, and you know you can avoid it just by using your brain and not stepping on the rake, and avoid entirely by just keeping your rakes contained, but a rake vendor comes to you saying that instead they have built a new rake that they swear won't whack you in the face even if you leave it right in your walking path, do you trust them?
1: https://www.youtube.com/watch?v=ouau9SVVrBA
reply