I think Mozilla’s controversial “Deplatforming is not enough” lays out a better strategy for “the algorithm” problem to me.
Having big tech be able to work in secret to pick anf choose what content isnt allowed and then being super charged by the state to do it for them as well just doesnt sit well with me
Motivation is powetful influence on devolopment. The linux kernel is largley driven by UX and desire for technical excellence (there are ultier motive from some major factions but overall this is true and actions are judged publically as such).
Microsoft is, like most companies, driven by stockholder value creation.
One produces an enviroment in which cautious adoption of new tech is constant, a slow trickle for use where it seems most applicable.
The other demands that the perception of exclusive capital be created through vertical intergration with propritary IP and that the promise of cost reductions are underway. Aka Microslop trying to add a buzz word to every IP (percieved capital creation) and promising massive layoffs.
The better setup generate “semantic embeddings” that try to map how data stored relate to each other (by mapping how to it related within in its own weights and biases). That and knowledge graph look ups in which the links between different articles of data are evaluated in the same way.
The very expensive LLM portion really do just give rough aproximations of information language in that setup
Huh, yeah. 50% cost in tokens could be 1 very expensive token or 1 million tokens. Who knows?
Im.very into “AsCode” and pretty comfortable with vi bindings. So the two extremes where i document (random notes with no structure needed and formally verified documentation ) it works for me.
I can preview the markdown, use vale rules to enforce style and vocab, do mermaid.js diagrams, link my UML to stuff, etc.
Then check into git to do version control or just to save it off local.
The silence of two hours ago lol
Chill brother. Like i get it, and FOSS advocates should lead with meaningful alternitives first imho, but there definitly seems to be some https://joplinapp.org/
I personally prefer vscodium and nvim myself for notes but that isnt a one for one comparison to obsidian (in either direction) imho
Its actually crazy how little is done for energy efficeny and EV preparedness for high denisty housing in the midwest US to me. There is just next to no incentive for most places because 1 they dont pay for electricty and 2 they dont have to tell potential tenets just how bad the bills tend to be in a place.
The EV preparedness is mostly just the lack of rent seeking potential vs any effort most rental companies are willing to put in.
Theoretically those bastards could be upselling power from meters they installed on the property and be making money from it, but that would require running a buisness with skilled and valued workers and not a constant revolving door of underpaid under trained employees.
“Hi im <girlfriends name>'s boyfriend. No no that is my legal name.” Behavoir
Right! GraphRAG, VectorDBs, larger context windows, MCP servers, “tools”, “skills”, and now “using the cli as a user” still havent really solved some of the inherent flaws of even the latest frontier models. SLMs and fining tuning gives me hope on addressing the learning part of machine learning. At least a little more
How have tests gone so far?
also cool concept, you can actually get eBPF XDP to compile to FPGA on some smart NICs even further pushing it away from the core system if it works!
Fair. I guess all air borne thibg suffer from being a potential energy battery (they fall) and have to be made much lighter to be reasonably energy effecrive.
Honstly RC car is probally better most of the time
That is a runtime that some flatpaks use as well.
Gnome and KDE as projects are a bunch of things, from login managers, to compositers, desktop UIs, and user application (like Gnome “System Monitor” or KDEs “Plasma System Monitor”).
You can actually mix and match some pieces and they just work, but especially the user apps because both teams put in work for interoperabilty or Freedesktop standarization.
So you can have an app that uses KDEs shared libs and an app that uses the Gnome projects libs on the same issue with rarely any issue. Even more so with flatpak since the all of the files those apps see are in what is called an overlay filesystem, so your kde apps get a layer of files jist for KDE apps to build off of and gnome apps get a layer of files just for them to use. In flatpak these are called runtimes. That is what is being updated here.
Tbh different tolerences. Drones can be smaller and can fly but also more ways to fail. I bet these (i mean for the price they better be) are more reliable. Even if its windy, snowy, or raining.
Honestly good on them. I hope they succeed and bring the project back to life.
I have my shared data on Longhorn, so for services that’s just longhorn as a PVC on rke2(k8s) and for clients I expose the NFS for mounts from a longhorn PVC to them to mount to.
Microsoft’s push to make Copilot a kind of AI medical middleman—especially through the newly announced Copilot Health—raises a real tension: the company is loudly promoting a Secure by Design philosophy, but the sensitivity of health data means the bar is far higher than a general security promise. The short version is that Secure by Design is necessary, but nowhere near sufficient for something that sits between you, your clinicians, your medical records, and your wearables.
Its crazy that this is the only sane option mentioned in this thread. Like its not a reasonable justification because the insane low trust situatuin we are in and no attempts to mitigate risks being heavily advertised first.