Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
barrkel
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
271.
▲
by
barrkel
1y ago
I think it's a little bit perverse to pile a lot of punishment on the OOW when it seems pretty clear that it's a training and experience problem, combined with years of efforts to cut crew sizes. The ship is lean crewed and relies
272.
▲
by
barrkel
1y ago
I'm afraid that this is a case of someone imagining that there are Platonic ideal concepts that don't evolve over time, that programs are perfectible. But people are not immortal and everything is always changing. I almost burst o
273.
▲
by
barrkel
1y ago
Ads are agents in a zero-sum war for your attention. If you have focus, if you're not aimlessly wandering around, ads try to distract you. They attack your personal agency and sovereignty, trying to divert you from what you were doing,
274.
▲
by
barrkel
1y ago
What is more common in many European countries is a check of a national debt register, to see if you're a known bad debtor. This is distinct from a credit check, because it's not something you need to build up; just something you
275.
▲
by
barrkel
1y ago
Once you start hosting images, you open up a whole bucket of worms for moderation, separate from handling storage, bandwidth etc. I can see plenty of rationale for deferring that, knowing that you can always add it later with an easier to u
276.
▲
by
barrkel
1y ago
The privacy policy references the privacy center for the conditions. They don't actually need to change the terms to retain and use data within the 5 year window. They can update the privacy center instead.
277.
▲
by
barrkel
1y ago
No, what makes a cache a cache is invalidation. A cache is stale data. It's a latent out of date calculation. It's misinformation that risks surviving until it lies to the user.
278.
▲
by
barrkel
1y ago
Caches suck because invalidation needs to be sprinkled all over the place in what is often an abstraction-violating way. Then there's memoization, often a hack for an algorithm problem. I once "solved" a huge performance prob
279.
▲
by
barrkel
1y ago
What if, instead of a bank account, it's FooSystemFrobnicationPreparer? Something which is necessary today, but probably should be refactored within the next year or two? Maybe FooSystem will be redesigned to take different inputs,mayb
280.
▲
by
barrkel
1y ago
I don't agree. When you implement some new code and want to cover it in testing, you have a choice where you call into the system from your tests. Calling at the lowest level is not always the right choice, but it's the way the te
281.
▲
by
barrkel
1y ago
Yes. When testing a component, use its public API, inject real implementations where you can, and use fakes where it's too expensive. Don't use mocks, don't test interfaces that have very complex setups to invoke, if possible
282.
▲
by
barrkel
1y ago
There is no marginal cost for training, just like there's no marginal cost for software. This is why you don't generally use unit economics for analyzing software company breakeven.
283.
▲
by
barrkel
1y ago
Summarization, very basic tool use, without needing to go across the internet and back, and zero cost because of edge compute.
284.
▲
by
barrkel
1y ago
Everything the agent has to read to pick out stuff costs $.
285.
▲
by
barrkel
1y ago
Ruby is slow, but it is however faster than Python last I checked, with like for like code. Python gives you other interesting ways of going fast (a lot of high performance plugins for numerics, Cython, and so on), while Ruby is a higher le
286.
▲
by
barrkel
1y ago
If you're doing OLAP, you probably want dimensions, measures and operators that operate on time aggregations and shifts. You want rollups and drill downs along multiple axes, with subtotals and probably pivots. SQL isn't wholly ad
287.
▲
by
barrkel
1y ago
CH is better for analytics, where append only is the normal mode of operation, but I've used it in the past as an index. Store a copy of data in Clickhouse and use its vectorized columnar operations for ad hoc queries (the kind where i
288.
▲
by
barrkel
1y ago
It's the other way around: the speedo can overestimate your speed but not underestimate it. If you follow the limits with an overestimating speedo, you drive under the limit. With an underestimating speedo, you end up over.
289.
▲
by
barrkel
1y ago
Done right, a more powerful engine needs uprating of a bunch of other components - clutch, brakes, suspension, tyre profile, cooling, exhaust, air intake and so on. Almost certainly not a big deal with a 10% bump, but if it was a big bump,
290.
▲
by
barrkel
1y ago
Alas, not true. It would be easier to predict progress if so.
291.
▲
by
barrkel
1y ago
Jules works in a VM, asynchronously, on a separate checkout of the code. Gemini CLI works synchronously with the user (unless you YOLO) and in your own directory on your own machine on your own checkout. Two different modalities.
292.
▲
by
barrkel
1y ago
But in practice they usually rely on attestation by an approved vendor, and the vendor won't let you control your private key, so they'll leverage it for lock-in.
293.
▲
by
barrkel
1y ago
It's not wrong. Putin felt free to attack Ukraine because he wasn't scared of the consequences. You need psychopaths to make a coldly rational decision not to mess with you, in international relations, and you do that by being pre
294.
▲
by
barrkel
1y ago
Cordura is the brand name for the usual abrasion resistant material in synthetic jackets, and it is a nylon.
295.
▲
by
barrkel
1y ago
I understand. However, I interpret your statement as coming from your position: as a moderator, with a job to do as a curator of a community. How does the community learn what to flag? It's not about piling on. You can fix disruptive
296.
▲
by
barrkel
1y ago
I understand where you're coming from, but there's also value in setting norms publicly, especially at times like these, at inflection points. The loud point and stare can have a bigger effect on behavior than the quiet disappeara
297.
▲
by
barrkel
1y ago
If you want to hide the LLM tells better, you need more consistent combing. E.g. "hasn’t", "they’ve", "wouldn’t", "“natural” and “synthetic”" all use curly quotes, but other parts use straight quotes.
298.
▲
by
barrkel
1y ago
[flagged]
299.
▲
by
barrkel
1y ago
Life expectancy in the past was heavily depressed by high child and maternal mortality. For those who survived into adulthood (say, 20), dying in one's 30s or 40s was not the norm. Most lived into at least their 50s, and many reached t
300.
▲
by
barrkel
1y ago
Tan is someone who can't handle disagreement or criticism. It likely leads him to live in an information bubble.
More ›