Kubuxu

It might be faster to burn the compute instead of having to fetch the KV-cache over the network from an SSD.

Pipes and underground pipes end up being better due to productivity bonus of the foundry.

If it is 30sqm local store, then get to know the owner and ask them to stock it for you.

This happens because the original publisher signed a deal for the commercial music with a limited time. So when that time expired, any further releases had to remove the music.

Novelty is a significant driving factor here. Gen Thumbnails are eye catching because they are new and viewers haven’t adapted to skimming over them like they do with everything else.

Thumbnails, same as everything that is supposed to attract attention, have styles that change over time due to saturation and fatigue.

AOO became a thing after LibreOffice essentially won. Before that it was OpenOffice.org owned by oracle, forked into LibreOffice in 2010, with LibreOffice getting significant traction and Oracle throwing in the towel and donating OO to Apache Foundation in 2011.

From the article, it seems they using a clause allowing them to terminate the contract after transfer of ownership.

This clause probably has a expiry date and they are saying “we are trying not to be jerks here by waiting until the last second to trigger it”.

As long as you are not running a paid bounty program. Otherwise now you are getting 40 per day.

It doesn’t make sense to do that, hardware deprecates in 5years (it used to be 3) and is currently very expensive. So by using this approach you double the average cost of hardware per unit of compute completed.

If there wasn’t such a hardware squeeze and electricity was a higher percentage of the cost, yes, it would have made sense.

There are rumours that Polish GROM has been taking “vacations” in Ukraine. As in “we cannot officially approve your presence there but we won’t stop you either”, from the government.

Because Shahed 131/136 are large, way larger than surveillance drone/quadcopter. We are talking 2-3m wingspans and 130-200kg launch mass.

I use SQLite history to have context aware line suggestion in my shell. So the command that gets ghost displayed (and can be expanded with right arrow) depends on the current working directory, its parents and recency. This is fast enough using SQLite to just dump all commands into it and query on each typed letter.

The DB is at this point 10years old. I use it to evaluate my work habits and patterns. More recently I have started transitioning to a different shell and I used it to move only the functions and aliases I still use, instead of the full collection.

Microchip also wants money for license to their design suite.

> So you read an article "China disappears random person" and then at the end you click a button and bet whether that person is alive or dead or whatever.

Congratulations, you just created a murder market. Create “Is person X dead?”, and put a large limit order buying No. Anyone with means can capitalise on it by fulfilling the physical obligation of the market and buying up the limit order. Don’t create the murder market.

---

It would be a stupid thing to do given the trail, bit it gives the person ordering it sort of plausible deniability.

I have TDD flow working as a part of my tasks structuring and then task completion. There are separate tasks for making the tests and for implementing. The agent which implements is told to pick up only the first available task, which will be “write tests task”, it reliably does so. I just needed to add how it should mark tests as skipped because it’s been conflicting with quality gates.

That somewhat existed in servers in the past, my R720xd had RAM mirroring mode. IDK if it used it for reducing latency, but you could take out a stick and the server would continue running as normal and report an alarm in iDRAC.

Yeah, after two paragraphs, my eyes started skimming it due to low signal-to-noise ratio.

Just try using Claude with API for an hour and you will see that the subscriptions are definitely not profitable (unless they percent off “partying but dormant” is very high).

Reasoning, in majority of cases, is pruned at each conversation turn.

A paper on the same topic: On the Expressiveness of Softmax Attention: A Recurrent Neural Network Perspective, Gabriel Mongaras, Eric C. Larson, https://arxiv.org/abs/2507.23632

Video presentation if someone prefers it: https://www.youtube.com/watch?v=PN3nYBowSvM

Linear attention is a first-degree approximation of Softmax attention, and model performance gets better as you increase the degree of the Taylor approximation.

I'm thinking about adapting an existing model to Taylor-approximated attention. I think it should be possible with some model surgery and rehabilitation training.

Definitely agreed, that is what I do as well. At that point you have good understanding of that code, which is in contrast to what the post I responded suggests.

I don’t see how I would feel comfortable pushing the current output of LLMs into high-stakes production (think SLAs, SRE).

Understanding of the code in these situation is more important than the code/feature existing.

And DEF is ~33% urea in water.

I've done that in the past (8+ years ago) with raw IP sockets.

For 8+ bays you just need a SAS HBA card and one free PCI-E slot. Not to mention that many motherboards will have 6+ SATA ports already.

If anything, 2nd hand AMD gaming rigs make more sense than old servers. I say that as someone with always off r720xd at home due to noise and heat. It was fun when I bought it during winter years ago, until summer came.

Traditional CMOS D-flipflop takes 16 transistors, SRAM cell takes 6 (two looped inverters and two data-line selection transistors).

The capacitor in DRAM is usually realised as an enlarged gate of a transistor AFAIK.

It is always PWM under the hood, the question is, how much was spent (or not) on the filtering network out of the PWM. Is it closer to buck converter or is it straight up flicker at the output.

Nobody runs 4096 samples per pixel. In many cases 100-200 (or even less with denoising) are enough. You might run up to low-1000 if you want to resolve caustics.

IIRC reordering additions can cause the result to change which makes auto-vectorisation tricky.

You can think of closing the channel as sending a message “there will be no further messages”, the panic on write is enforcement of that contract.

Additionally the safe way to use closing of a channel is the writer closing it. If you have multiple writers, you have to either synchronise them, or don’t close the channel.