In our Sysbench tests, switching WAL commits from fsync to fdatasync and prewriting zero-filled journal regions improved Dolt’s write performance by 10–40%. Aaron explains why the change works, where it doesn’t, and some of the tradeoffs: dolthub.com/blog/2026-10-13-…

everyone is asking us at @cxodev about calculating ROI on AI, here's what we're saying: 1. focus on engineering, it's 80%+ of your spend - if you're looking to improve cost efficiency its the big lever to move (so stop wasting time in other depts) 2 . model tests + evals are your friend. learn to run them for the most expensive parts of your business 3. spend time with your top spenders: they're either a) doing things your full team should do or b) an easy spot to save some $$$ 4. you're probably underspending relative to oppty all things considered 5. PR velocity is a good not great measure of eng ROI. you need to tie to actual product intiatives and use 6. if you're not sucking up session data + traces you can't really tell if your team is using AI on the right things 7. everyone is blowing through their 12 month budget in 4 months. you're not alone 8. aim for 2-10x improvements in ops efficiency. 30% is a rounding error. 9. time saved =/= ROI. what are you doing with that time? 10. you should be eliminating whole categories of human toil 11. nobody calculated the roi of email. some of this is simple work transformation 12. idk we still like tokenmaxxing 13. you will spike up and then it will be possible to pull down costs (but not output) - we see it all the time. intelligence is getting cheaper. 14. revenue per fully loaded employee is emerging as the north star guide. most leaders won't say that out loud to their teams. 15. spend capping your top users is dumb. they're the ones most likely to be highest impact if you really look. 16. sorry, but you just don't have that many fable-level problems. stop making it your defaut. what am i missing?

This is the most insulted I've ever been on Twitter in ~20 years, and that's saying something. Not to pull receipts, but I was the first multi-pint drinker on Cheeky Pint, and we spoke in detail about various beers on the episode itself.

A Possible Explanation: The US knew there was a threat to RAF Fairford - that's not in dispute here. The FPCON changes, armed patrol vehicles inside the base and concrete barricades all show that. It seems possible to me that perhaps these 5 lads (or 8-10 if the eyewitness farmer is correct) were not actually there to attack the base - perhaps (given the fake fuel vans) they were actually in the area trying to steal fuel from local vehicles. The farmer sees it, calls police. Police (who likely were aware of the USAF concerns) swarm into the area having perhaps not expected the "attackers" so soon. They manage to arrest 5 suspects and treat the 3 vans with every necessary precaution (as they should) - evacuating the neighbourhood, bringing Army EOD in, etc. The males are initially arrested for Explosives Act offences (pending a safe way to check inside the vans) & go to custody. In the circumstances, further arrested for the Terrorism Act offence. However, after seizing whatever evidence they could (and examining the vans), it looks like police may have determined these aren't the attackers - as I say, potentially fuel thieves in the wrong place at the wrong time? It's worth noting that if the terrorism act offences were looking likely, police could keep them detained for 48hrs initially before requiring a court to approve a longer period of time. Alternatively under normal arrest powers contained within the Police and Criminal Evidence Act 1984, they can be held for 24hrs initially (extended to 36hrs under authority of a Police Superintendent if specific requirements are met), and up to 72hrs if a further extension is authorised by a court. Now clearly if they were arrested in the early hours of Sunday morning and conveyed either to a local custody, or a dedicated counter-terrorism custody (likely closest in London), the fact they've been bailed today suggests they've spent at most 24-30hrs in custody before evidential review officers and the Crown Prosecution Service have determined there is insufficient evidence to charge them with an offence right now. So they bail them with conditions not to contact each other as is normal procedure for vehicle interference/theft offences like that. But police need to download and analyse phones, and anything potentially seized from home addresses in order to ensure no link to the Iranian government or IRGC. Meanwhile, did the actual planned attack get foiled inadvertently? Perhaps? Trump told the world that the US had been monitoring the "attackers" for some time - I'd suggest that if these lads genuinely were just in the wrong place at the wrong time, the actual suspects are now staying hidden until this all blows over a bit. Either way, we don't have all the facts and that's part of any criminal or CT investigation - so yes, it sounds bad and arguably CTP could have done a better job of explaining it to the public, but they have to play their cards close to their chest right now in case my guesswork here is right and the suspects are still out there somewhere.

We believe superintelligence will create significant new opportunities for all people and businesses. Meta already serves billions of people at scale and helps hundreds of millions of businesses reach customers. Today we are starting the next major pillar of our business, Meta Enterprise Platform, to help businesses use AI to grow and transform in new ways as well.

We rewrote an internal thing from Zig to JavaScript and it got faster and more reliable. The obvious conclusion is that language doesn’t matter. Rewriting does. You should be rewriting everything all the time.