Mama told me not to come.
She said, that ain’t the way to have fun.
Could it be correlated to Chinese New Year? I could see a ton of people in China suddenly playing and then not playing if most people get time off at the same time.
Does it happen every year around this time?
Alternative perspective, it could be related to the Great Firewall of a China, perhaps they made some rule adjustments and data came through that doesn’t normally come through.
egress costs, which is their true and much sneakier vendor lock in trap.
Absolutely. That’s basically Oracle’a db strategy.
Things like this are why I’ll never use AWS, even if I get to a scale where it makes sense. I value the ability to switch to a different provider or self-host with my own hardware.
the only answer is government regulation
Ideally the market is competitive enough that regulation isn’t needed. But maybe that ship has sailed.
I agree with regulations like Net Neutrality, so I guess it would depend on how it’s worded. I’m just worried massive players like AWS would find ways to abuse any regulations we try to make to exclude others.
But yeah, I don’t pitch switching at work, because I’m not in charge of infra or really involved with it at all. I’m a SWE, not a devOPs or IT tech, so if I’m touching anything in Cloudwatch other than looking at logs, something has gone horribly wrong.
Just try mounting it, if you get bucked, try again.
We use AWS at work, and the “cutting costs” thing seems largely a way to further lock-in customers. They want you to build around their tools so the switching cost is high enough to not be worthwhile. Then again, I don’t work directly with billing (I’m a SWE, not in OPs), but what I’ve seen looks a lot higher than I would’ve guessed.
Idk, maybe it’s reasonable at scale, but it seems to get really expensive really fast.
I got an ASRock x370 board and it had fewer problems than other first gen Ryzen boards, was one of the first to support 5000 series CPUs, and it’s still working well in my NAS with no dead ports or anything.
They used to be one of the lowest quality boards, but they’ve earned my respect. My current motherboard (b550) is also ASRock and the only problem I had was the WiFi chip sucking on Linux (known issue, I replaced just the WiFi module and everything is golden). My wife has a Gigabyte board and it had some minor issues.
So yeah, ASRock is on my good list for now.
It’s a great article, actually click through and read it if you haven’t already.
My favorite example of truly effortless communication is a memory I have of my grandparents. At the breakfast table, my grandmother never had to ask for the butter – my grandfather always seemed to pass it to her automatically, because after 50+ years of marriage he just sensed that she was about to ask for it. It was like they were communicating telepathically.
That is the type of relationship I want to have with my computer!
The author’s point is that natural language is a slow way to communicate, and it’s not even our preferred way, so why are we pushing so hard for it?
One of the best productivity tools for me is my CLI shell, which predicts what I’m about to type based on what I’ve done in the past. There’s no AI here, just simple history search. It turns out i do the same thing a lot.
None of this is to say that LLMs aren’t great. I love LLMs. I use them all the time. In fact, I wrote this very essay with the help of an LLM.
The author argues that LLMs are an augmentation to existing tools, not a replacement. Just like the mouse didn’t replace the keyboard (my example), LLMs won’t replace existing workflows, it’ll merely help in the knowledge retrieval stage.
For this future to become an actual reality, AI needs to work at the OS level. It’s not meant to be an interface for a single tool, but an interface across tools.
This is where I partially disagree.
Yes, I think some level of AI makes sense at the OS layer, but its function should be to find the right tool, not to be a tool. For example, “open my budget” would know from context which file that is (family, company, client, etc), which program (GNUCash, Excel, or a URL in a browser), and then pass on context to the app-specific AI, which would know which part to open and be ready for context-relevant questions (is it payday? Was I just looking at concert tickets? Is someone’s birthday coming up?).
But even then, the usefulness of a system-wide AI is pretty limited. Most people can efficiently navigate to what they want. Indexes work well to find files (and full text search is feasible), file extensions work well to open the right application, and applications remembering what they were last doing is usually sufficient.
So I see it as more of an accessibility feature at the system level instead of an actual, useful system in itself. However, I really like the idea of different models passing context in some standard way to each other so I can seamlessly move between apps.
But I absolutely agree with the main point here: AI should be seen as an add-on, not a replacement.
The thing is, I don’t really care about Android apps, and honestly supporting them probably adds a bunch of limitations since they have a lot of expectations on the system.
I just want an immutable base system w/ flatpaks, a basic dialer, a robust SMS/MMS app, Firefox, and good enough battery life (15 hours w/ moderate screen on time). Basically, openSUSE Aeon or Fedora Silverblue with phone-specific apps.
I’m happy to help port the various software I want to use, but I need the phone to work as a phone first.
Absolutely! The controls might suck and regular phone features might be iffy, but you could totally run the Java version of Minecraft if you wanted. No guarantees about performance though.
And we’ll keep saying it until we get there.
It’s also less interesting.
Using regular Linux means you can do a ton of stuff you currently can’t on Android:
Android is already FOSS, and you can get phones with minimal stuff on top of the FOSS core. That’s cool I guess, and I use one such distro (GrapheneOS), but it’s still Android at the end of the day. I want something different, but I still want basic phone stuff to work (calls, SMS, MMS, camera, etc).
I’m okay with an older phone, I just want basic features to work consistently and well. Maybe support a newer phone every 5 years or so to provide an upgrade path.
Basically, I’m okay with the GrapheneOS strategy of sticking to one product line.
Hmm, I seem to recall Trump being a pedophile.
Now you can keep out LPL for 30 seconds instead of the usual 10-15!
The part about electricity is pretty short, just a few bullet points and a couple paragraphs. If it was video form, I could cover it in 2-3 min complete with visuals.
I don’t know why this video is so long.
No, it’s lack of experience. When I was a junior dev, I had a hard enough time understanding how things worked, much less understanding how they could be compromised by an attacker.
Junior devs need senior devs to learn that kind of stuff.
Nah, the video is pretty straightforward, but it’s presented in a way that most people would lose interest unless they’re actually into the subject matter. I see three problems with the video:
Power vs energy is fairly simple with a good explanation. Power is simply the speed at which energy gets used up. For example:
That’s extremely intuitive. All a regular person needs to know is that simple concept, plus a way to measure it (the Kill-a-watt example). Boom, 5-10 min video.
But the talking head made it way more complicated by starting with gas. That’s just belabouring the point that you can increase or decrease power, which is already intuitive with batteries or things we can see (wood) or feel (tiredness).
They could then segue into gas, once the power vs energy issue is established. A can of butane is like a battery, and the valve (e.g. screen brightness, game vs texting) controls how quickly it’s used. We can compare gas and electricity directly because electricity can be turned directly into heat, just like gas can.
And then you segue into heat pumps. Basically explain how your AC/fridge works (i.e. moves heat instead of creating “cold”). Make a demo where you move heat vs create heat and show how much energy is used. As in, heat a room from 72F or whatever to 90F, one using a heat pump and the other using a space heater. Show how temps compare on both sides of the heat pump vs space heater (other rooms shouldn’t change w/ space heater). Then use that to show a real-world example of a house that swapped from furnace to heat pump to really drive the point home that moving heat is more efficient than creating it.
price at enterprise scale
Really? I thought that’s where big cloud services fleece customers the hardest… We use AWS at work, and I’m always surprised when I ask our devOPs how much we’re paying.
My understanding is they’re selling the “time is money” angle, where things work together well so you spend less time getting stuff set up.
But why? There are already a lot of great services based in Europe. For example, Hetzner and OVH. Their product offerings aren’t exactly 1:1 w/ those big three, but they have a lot of great tools, and you can get pretty far w/ a DIY approach, you just need to hire some OPs people to manage things. Hetzner even has S3-compatible storage.
I get that there’s a lot of interesting abstractions w/ places like AWS, but I’m also of the opinion that a lot of it is unnecessary and just adds cost. Learn to orchestrate things properly and build some tooling to utilize the APIs these cloud services provide, and you can achieve the same thing for less cost.
Nvidia is fine and has been for decades. I used Nvidia exclusively on Linux for ~10 years and it was fine. AMD is a better experience on rolling releases and for keeping up with Linux changes (e.g. Wayland), but Nvidia has always been fine, you just might be stuck on X or obeying something and miss out on some features like having two monitors with different refresh rates (IIRC that was only possible on Wayland).