Not entirely exclusive. There are lots of smart people who still support Trump in 2026. The problem is that their objectives are extremely malaligned with the rest of society.

I don’t know… Have you provisioned Cisco equipment before? It feels like the quality can only go up.

(But no, obviously this will end terribly unless they’re incredibly careful. It’s possible, but unlikely.)

Note: It’s unclear whether this is an adequate mitigation.

Those people with inflation fetishes must be having the time of their lives.

Thanks. And yeah, the wrappers start the service if it isn’t already up, then it either times out after some period when the VM isn’t in use and shuts down, or I close it manually if I want to free up the resources. But it’s pretty lightweight.

Make friends, explore hobbies, be eager to learn

The dude in managing to slur their speech in text and somehow all you can see is capitalism; typical ML user /s

Huh. What an unexpectedly damning indictment of the extraction-oriented enshittification market aka capitalism

Really depends on how much headroom you need and your system ram and the PCIe / vram speed because I got a friend with a pcie5 16gb gpu and 64gb of system ram that’s getting slightly better performance than me using the same setup, despite half the vram.

If you want speed and not a ton of intelligence, the 3.6 30B A3B runs fast. If you need smarts, then either a qwen 3.8 27b (probably a thinking cap finetune) or try flash next (swift finetune) in unbiased strata if you don’t mind taking a risk and burning some bandwidth on a 75-85GB model that might not run great. In either case, be sure to use MTP and the model card inference params

Aww and it was so obvious too

Democrats are a bit like Internet Explorer. Slow, painfully behind, only listening to the people’s demands when a court orders it to, constantly forcing itself on you when you don’t want it, and sometimes it’s all you got.

Just started testing this out, also upgrading to the bigger iq3_s; working well so far thanks (tho only been like an hr…)

I built my custom solution with a shared systemd qemu VM service to run harnesses isolated from my host that mount specific directories and run via wrappers that start and automatically SSH into them, as well as native variants that run on the host machine. (Plus args for quick selection and feature toggling e.g. “cx sol high memory” as well as per-harness hooks for my agnostic skill-like text injection system, memory tools, optimized system prompts, etc)

All of this is done in my nix flake (but can’t share here)

For Deepseek itself, I’m just using Moraxyc/deepseek-harness.nix for the package, and using its webui, with wrappers to integrate it with my vm/native dichotomy system.

I haven’t done any deep tooling integration or prompt rewriting with dsh like I have with codex and claude code. It’s still pretty early days for me using it and I prefer terminal apps in a multiplexer like herdr. But it doesn’t seem to need it so much either: sonnet eating 35k tokens for a “hello” out of the box was egregious to me, but dsh seems to be more light so I haven’t felt the need, and I don’t do as much with it and haven’t used it for very long yet so I haven’t felt the need to customize it as much as I have the other harnesses… Which is a bit ironic given that its customization and flexibility are pretty much its main selling points.

it carries around a lot of baggage from the last 40 years.

Cppfront2 is an interesting experiment that looks at an approach to solve that problem without a hard break. While my C nostalgia is offended by the syntax, I’m quite fond of it. Bjarne’s pet topic of ‘profiles’ is another but hasn’t landed much yet.

Well, I think a caveat to that is publishing source (no prebuilt release) as unvetted slop, so everybody knows. Though in this case, it was strictly for personal use.

But yes. Anything I release proper, for general consumption, is owned by me. And professionally, I never own even a line of code (or comment) that I don’t understand or wouldn’t have written myself. I may ask for a bunch of audits and slop tests for insights, but I’ll keep those offline as “extra” unless it’s something on par with I would have written. I’ll never hallucinate important code, and anything i get paid for or contribute to OSS is important. Personal game mods and small utilities, not so much.

That really hurts to read. Have you programmed with any spec newer than c++11? Best-practices have changed a lot. It isn’t “c with classes” anymore. Yes, most of the ancient code will work; but there are better, more efficient, and more ergonomic alternatives now.

The problem with QT is that everything becomes a “QT app” when you start writing with QT. You’ll end up using more of their types than the standard library, and getting a clean front-end / back-end separation is more difficult.

Anyways, C++ isn’t just a superset of C; they’re surprisingly different, and the best modern approaches are extremely so. That advice may have been true before 2014 and a bit until 2017, but it certainly isn’t now.

Over 15 years of low level c and modern c++ programming for resource-constrained systems; tutored it too. Feel free to DM if you have questions or get confused about anything.

Haven’t used opencode yet so can’t compare. Dsh is pretty nice and customizable, and its compaction seems to work well. Qwen doesn’t seem to like keeping the todo up to date and can’t easily differentiate harness vs user instructions. Not sure which is at fault.

So backwards. We all know MAGA cultists would do the sucking.