How uv Works Under the Hood | Noos - Where Thought, Code, and Craft Converge

A thorough walkthrough of uv's internals: the Rust crate architecture, what uv init actually does on disk, the two-thread resolver design, how PubGrub's CDCL algorithm works, batch prefetching, the forking resolver, and why these Rust-specific patterns make it 10–100x faster than pip.

Noos - Where Thought, Code, and Craft Converge
futures are not down much, eventually there will be some sort of deal and closure within weeks, in ukraine it is going to take them all summer to get 25% back, it drags into next year, in a spy vs spy game don't expect too much, doctrine is shifting a bit, the front line is 40 miles wide now, the autonomous swarms will make a difference, will these evolve into superpacks, maybe, since it is sort of supply chain and logistics war at the core.... the tide is turning - see what happens this week. ai could help win the war but it is just math/algorithms, not biorhythms #patterns #killchains #how the west was won - ai #ai: more dangerous than a global nuclear war?
Weekend Side Note
If a hypothetical investment fund of significant size, say $180B, were to badly underperform for 2025, the results have serious implications.
In Canada, the TSX performed an astoundingly rare return of 28.5%. A well- managed fund performing at the rate of the market should expect a return of $50B resulting in a fund sized around $230B at year end.
If a hypothetical, poorly managed investment fund gained, say, a paltry 7.5% on such a generously giving market year, they should get a return of around $14B and final fund total of $194B at year end.
Counted in billions, 14 can appear a big number unless weighed against its absence of return. The poorly managed fund has actually incurred a short-term “loss of gains” in a rarely generous year of around $35B.
In the long term, over say the next ten years, that loss of $35B increases, at a yearly return of the initial paltry 7.5% (normally good for an underperforming year), to roughly double or $70B.
Fortunately, if the hypothetical fund continues under its current mismanagement, the ten-year 7.5% return is unlikely, thereby reducing the “missed out on” $70B to around $40B as best, or in this case worst, result from an extreme hypothesis of missing funds during one calendar year. Such “loss of gains” exaggerated annually accumulate to massive losses of squandered opportunity over decades.
In the real world, it is merely $14B thanks to the generous “hard to lose” market year, and the loss of opportunity goes unchecked while the pennies get counted.
#how #money #works
How To Write Unmaintainable Code

How To Write Unmaintainable Code in Java

How to turn anything into a router

I don’t like to cover “current events” very much, but the American government just revealed a truly bewildering policy effectively banning import of new consumer router models. This is ridiculous for many reasons, but if this does indeed come to pass it may be beneficial to learn how to “homebrew” a router. Fortunately, you can make a router out of basically anything resembling a computer. I’ve used a linux powered mini-pc as my own router for many years, and have posted a few times before about how to make linux routers and firewalls in that time.

Noah Bailey

Remember that #paper and #pencil are also technologies. They are the result of a process. The big difference, supported by studies, is that writing by hand (and drawing) activates areas of the brain and develop skills that are not activated by just typing on a computer keyboard. And many of you, like me, have experience with the #typewriter as an intermediate step.

With #genAI, the risk is to underdevelop even more thinking areas of the brain.
Throughout the history of humanity, we have observed that the human mind shapes #tech, and vice versa, tech influences the #mind.
In a particular moment of history, #Luddites were destroying tech.

But there are more brilliant ways to face the issue. People should understand that #AI is a tool. Not the new Delphic oracle, not a devil.
The issues are: for #what it is used, #how many resources are needed (e.g., water), #which data are used for the training, and #who handles data. Every kind of tech should #empower people, not making creativity obsolete.

Oh, and if you stop eating #meat, you are sparing a lot of #water as well.

#regulateAI