6,621 karma · joined July 5, 2023
It's great having tools like Facet around to add runtime type information in. I wish TypeScript was seeing progress but there's seemingly no progress or traction in 5 years+, just some early experiments which died out.
> The original design was conceived by two Lockheed Skunk Works engineers who had prior race car engine knowledge. In the mid-1970's, the engines of the start carts was changed to Chevrolet LS-7 454's. The Chevrolet big block engines developed 465 horsepower each. Even so, for many years they were still referred to as the Buick's.
https://www.thesr71blackbird.com/Aircraft/Engines/starting-t...
And eventually the start carts themselves got replaced.
> In the '80's the AG-330 Start Carts were mothballed and in there place came a Pneumatic Air starting system for the SR-71's. It is true that the start carts were hard to maneuver around. Logistically, a recovery of a Blackbird at another base could be done easier with Pneumatic Air.
So, what? The world just isn't allowed to write secure code? Not without permission? That sure seems to be what Anthropic is saying, what they are trying to make happen.
The smoking gun is how much slower than Sol 6 this is. It's not a retrain.
(I did use some CC for Fable when it came out, and it was... ok. Not the worst thing ever.)
As someone paying for my own usage, the transition from a time of plenty, of eager happy use of my subscription, to begrudingly pulling the token-dispenser handle from the token machine, knowing each query costs me: this is a majorly sad transition to see happen. I thought we'd get at least a year of subscriptions being a joyous amazing thing, but no, already the hand that giveth doth taketh away.
Fun facts: the Trans-Bay Cable is credited by wikipedia at least as the first MMC implementation. Started ~2003, permitted 2007, completed 2010, the cable is 53 miles, and runs at 230kV, providing reliable power to SF; a response to a critical 1998 power outage when two local power plants shut down (since decomissioned for good) and the grid failed to carry the load. https://en.wikipedia.org/wiki/Trans_Bay_Cable
Coincidentaly, the original patent expired in 2003, so uh, interest may have started rising recently. None the less, thanks Rainer Marquardt and Anton Lesnicar for your MMC invention. https://ieeexplore.ieee.org/document/1304403
I could be wrong but I believe the QWidget's that make up nearly every KDE app had DCOP built in to them.
FDP is a very minimal addition of control. But yes it is another box to ticket, is another place to upcharge. Yet still, we're only just seeing mainstream products emerge. Kioxia's CM10 for example. https://www.techpowerup.com/351218/kioxia-introduces-first-p...
I'm hoping that the need is great enough to break the industry control. I think for a while the market felt relatively well enough served such that it was unclear whether clear wins would actually result in customers. With AI need for speed, I can definitely imagine incredibly crazy CXL controllers or what not, that allow low latency acces to many many open channel flash systems. Wouldn't that be a thing.
Again though, FDP is such a ridiculously tiny add, and it helps SSDs so much for so many use cases. I really hope it becomes an expectation, not a feature, sooner rather than latter.
Edit: happy to see a new group has shown up asking for open channel flash, Open Flash Platform. https://openflashplatform.org/
I do really appreciate Klassik here though in general. There's so much nostalgia work for recreating or preserving various old systems. It's a noble tribute to see classic linux desktops also paid that honor.
There's only a small little mention, if we are lucky, on the couple drives that have it (expensive enterprise flagships). It should be a regular sticking point, whether it's there or not. Without pressure it's not going to get regularly available, it feels like.
FDP is so simple. Declare a number for what pool of data you want to write into. Data of the same pool gets written to the same storage such that you can wipe it latter together. It has huge wins though against write amplification! Massive wins. For so close to free.
Some day I want to own a FDP drive. And then I can finally start using the tokio/io-uring support that I contributed! https://github.com/tokio-rs/io-uring/issues/380
The NVMe-KV is more radical. Still worth putting some pressure on, but your drive as KV, as object store, feels harder. Side note, really enjoyed this ceph nvme-kv offload post thing, my favorite tech write up in a while! https://ceph.io/en/news/blog/2026/for-whom-the-door-bell-tol...
Edit, follow-up:
DSRr in rust: https://github.com/krypticmouse/DSRs
I was charitably thinking perhaps maybe possibly being a fork doesn't give you access to the full suite of GitHub tools: iirc issues aren't at least by default available? Maybe there's a setting for that.
It felt like a little too big a topic to tackle in my post, so I ommited, but I definitely have concerns! Very weird.
This one appears to be a more active fork of Tobi's, https://github.com/tobi/walgit https://news.ycombinator.com/item?id=49420598
My comment listing 6 from 2 weeks ago: https://news.ycombinator.com/item?id=49763869
Nesbox is just an underlying part of a broader (but early) kit to allow a system to stream multiple remote desktops at once. https://github.com/nestrilabs/nestri
KDE krunner has plugins. But there is not afaik a big tent ownership system like Raycast. And krunner is only going to attract KDE users. Raycast is mainly mac, but also, there's Vicinae which runs on a lot more systems, and it can run >50% of the Raycast plugins without a sweat, is designed to be drop in compatible.
It's also absurdly easy to write Raycast extensions. It uses web platform technologies that "just work", that don't have ornery toolchains and need lots of dev libraries installed. With KRunner you can probably use whatever library you want to do whatever you want, which is good, but hard, where-as with Raycast there's a pretty comprehensive and powerful API that offers a ton, that is well defined, and that other plugins will show you the way on. https://developers.raycast.com/
Some backdrop, a thesis: tech used to be exciting because there was apparent motion, things changing. The dread hadn't set in, and I think that dread set in because tech shifted from close intimiate and touchable technologies we could touch, to far off systems running in some cloud data center/keep. Change was fun when we got to be on for the ride and feel some ownership but change happening to us was annoying.
So lately, it's hard for anyone to know what their preferences really are. We aren't trying a ton of human empowering things. The organizational systems Scott talks about here? We see some absolutely ferverously dedicated people enjoying Niri, enjoying new infinite space tiling systems. That's a rare example though! A couple people stumble their way into it. There's not that WinCustomize era where it's apparent people are active and busy and rebuilding, remixing their world on and on. People aren't in a receiving trying mode, because in part there's not a ton to try, and because there's no scene, people aren't kicking out new things either. The social poverty of the area & lack of things to look forward to mean people aren't looking, and people trying stuff are lucky to get traction, even when it's good.
This admits some of the "don't touch my stuff." Because we haven't been seeing as much upside to change. But also: it's pathetic! These peole have had their spirits crushed and have no peers out there who are having fun, so they can just wallow in their grumpiness. And we need to kind of ignore these users, because they don't know: they think they want not even faster horses but merely horses that stop breaking down or which start only turning left unless you pay premium. Let these naysayers and grumps alone. But more than anything: show the people you can reach good stuff! Activate interests! Try to restore a sense of motion & exciting change! Create more powerful, richer, more beautiful systems!
Agents help though. Change sucks when it gets in the way. The downsides though are lower than ever, if you have systems flexible enough to let agents work them. If something changes in a way you don't like, an agent may well be able to make quick work of it! Tell it, and it'll go fix it.
Last: the granularity of build it yourself systems is not great. Downloading Sway, a desktop shell atop it,then a launcher, then a host of other small utilities to cobble together is a lot of decision making to do, and most of these options for these roles have lots of competitors with little interoperability. The tools that makeup the desktop stand alone, small pieces, and it's up to you to pick each one and loosely couple it, with it's bespoke means of doing so. These patterns of assembly suck. There's significant money heading towards a new omakase system, and I don't think it's necessarily the authoritarianism/fascism per se, the desperation for authority to make sense of the world, but given how scattered the pieces are, I have some sympathy for why they want someone to say they've built the beautiful well hewn omakase thing. And that your agent can customize it.
One of my favorite examples for a system that does compose nicely is astronvim on neovim. It takes some well known very great core plugins for nvim, and builds a little zellij like mini-framework to nvim's tmux atop that: a visual leader key, with menu system builtin. It creates a framework for more to plug into. And then the community has lots and lots of plugins or packs of plugins that it takes from upstream plugins everywhere, and it packages them for astronvim. Hundreds if not thousands of plugins. Most require not that much extra to hook in nicely in to the framework, to make them more apparent, visual, seen in the broader system. The system composes capabilities into itself nicely, integrates. Where-as when you get a regular nvim plugin, it usually can assume either nothing or way too much about what kind of nvim it runs on. It has to define it's own world. I think astronvim is an incredible model for making common layers for assembly, and it's packs show something inbetween a single plugin and an app: it's a unit of software that aggregates a couple other modules, yet is shy of being a full app. That kind of composability of experiences is unique and special: not having to assemble every piece ourselves, and yet not having an entire massive system dropped on us. Figuring out how to deliver better units of capabilities, composable subsystems, is really missing on the desktop and is much needed.
I'd love to see an agentic OS where the OS and apps can both talk in terms of MCP things like resources, tools, skills. Yes these are primarily geared towards agents, but they are also acceptable machine-to-machine interfaces to script against too (so long as the type signature remains compatible).
Android's Intents feel related, similar-ish. It's a pity no one bit on PK's WebIntents. That could have been an interesting way to create abstractions for intents, for activities, for stuff happening on the computer. Notably the xopc actually had activities and a journal that let apps contribute meaningful intent to the OS as a whole. But mostly it just ended up as a history viewer there. Android mostly uses intents as a launcher. But this idea of having the activities of the system as available abstractions could go much further.
nothing wrong with it! looks fine! just: there should be other products with the nice core. i could really use a 10-14" tablet. i mean, after all, ROG Flow 13 with a Strix Halo was not even that late after launch! dense is doable!
Please sell >1 model of computer with X2 EE (elite extreme) please please please. Not that #1 was bad at all but more please! https://www.notebookcheck.net/Asus-Zenbook-A16-Laptop-Review...
I really hope it's like 2031 and we're swimming in way too many of these. Running great Linux. How sick would that be. (And X4 is running mainline great too).
This post, this title resounds true: your user agent is only your user agent if you two have freedom to work together, to improve your agency together. A fixed set of capabilities by a service provider that they offer you will always constraint and bound.
You can and should have a system that offers the real tamale, that you and your agent can extend improve the agency of kind of without limit. The Cloud agents and their fixed slate of what they do is just an ill compare.
That said I do think there is incredible value considering new scale out computing architectures that are hosted first, but general. Systems like Agent Substrate and Ax aren't exactly the general purpose system we know. But if they allow users to launch thousands of their own scripts to run ambient in a cloud, with good platform underneath: that will be a kind of phase change in computing, that makes abundant the ability to have your agencies/capabilities (the things you and your agents launch, make) more freely available. https://news.ycombinator.com/item?id=49780797 https://agentexecutor.io/
There is, as there always is, a huge dual. The prescriptive vs holistic technology set, of what are you being offered that's a hard cast thing, vs what is clay and bone you can lay freely. Note how work vs control technologies so closely abut's Ursala Franklin's prescriptive vs control: https://en.wikipedia.org/wiki/Ursula_Franklin#Holistic_and_p...
In these demon's blanket national insecurity, this is their chief enemy? Their chief rallying cry, what they put themselves in opposition to? The only politics they talk? Must be something there!
Quite the choice to say he doesn't deserve attention. Doing a public work to raise some much needed attention, to give people a basic heads-up what they are stumbling into.
Once again here Drew just letting the fishes know what particular brand of wet the water is.
But the system itself so conveniently allows for just a couple people to suppress and deny, to shut down topics, that it systematically gives quarter to the very worst, to the motivated bad actors, and the other lesser plagues that befall humanity.
I really wish there was some accounting, some visibility, some consequence for this abuse of the system, this steering away from hackeryly inquiry & examination of the world. The act of cloaking & hiding shouldnt be so easy to do from the shadows; HN should at least make you do such moderation acts in public.