Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
barrkel
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
34 ms
·
601.
▲
by
barrkel
3y ago
In Germany and Switzerland, washers and dryers are usually in the basement, otherwise in the bathroom. In the UK and Ireland, they're usually in the kitchen. Unsafe to generalize.
602.
▲
by
barrkel
3y ago
I live here and have rented vehicles across pretty much all brands and have never seen a car where the front passenger airbag can't be turned off. In fact there are ~identical regulatory stickers on the passenger sun shade warning that
603.
▲
by
barrkel
3y ago
McD apps for both Germany and Switzerland feature coupons. In Switzerland, physical coupons show up in my postbox. For whatever reason the coupons very seldom work for my meal preferences so I don't bother with them. In Spain the main
604.
▲
by
barrkel
3y ago
It's an old trope from Perl. The virtues of the programmer: laziness, impatience and hubris. https://wiki.c2.com/?LazinessImpatienceHubris
605.
▲
by
barrkel
3y ago
Russia is barely able to keep its planes in the air, and you think they're a competitor?
606.
▲
by
barrkel
3y ago
Luggage isn't more convenient on trains, and car rental often isn't super convenient either, if you need e.g. a 10 minute walk through city streets with all your bags to get to some multi-story car park.
607.
▲
by
barrkel
3y ago
Yes, this. If the form of a plausible answer is known it is likely to be invented. API method names, fields in structures, options to command lines, plausible inventions that have a known form. Similarly references of any kind really that h
608.
▲
by
barrkel
3y ago
You only need one recursive CTE typically and you can encode it into a view that you join against. It just sucks for performance when you have millions of nodes.
609.
▲
by
barrkel
3y ago
I found using recursive CTEs incredibly slow for assembling tree data from a normalized representation. It would make getting the results of a query at least d times slower for assembling the path of nodes d deep in the hierarchy. The upsid
610.
▲
by
barrkel
3y ago
This kind of encoding is usually efficient when the column is indexed because it can use prefix searches.
611.
▲
by
barrkel
3y ago
No. Think about it: how could the underlying distribution change the probability of a hit from a uniform random sample?
612.
▲
by
barrkel
3y ago
As long as the sampling isn't skewed, and is properly random and covers the whole space evenly, it will estimate cardinality correctly regardless of the underlying distribution. There is no way for clustering to alter the probability o
613.
▲
by
barrkel
3y ago
I'm increasingly finding that my phone is detected while I'm still moving it towards the scanner, and it complains that my phone moved too fast. Then I have to wait like a schmuck until something resets and I can do it again.
614.
▲
by
barrkel
3y ago
FWIW, I've found the VW group software for traffic sign recognition significantly more accurate than BMW, and I like the lane assist, it's far less intrusive than other implementations I've driven. (I'm not including lan
615.
▲
by
barrkel
3y ago
https://www.google.com/search?q=gradient+banding+4:2:0 There simply aren't enough bits of precision in the luma encoding for good gradient support most of the time, chroma fills the gaps, and chroma subsampling produce
616.
▲
by
barrkel
3y ago
The gradients on webp frequently look like video stills. Chroma subsampling reduces the density of available luminance approximations and the more heavily it's applied, the worse gradients look. High contrast high frequency details are
617.
▲
by
barrkel
3y ago
Oven cleaner is a better bet.
618.
▲
by
barrkel
3y ago
The difference between running llama.cpp main vs server + POST http request is fairly substantial but not earth shattering - like ~6s vs ~2s, for a few lines of completion, with 8GB VRAM models. I'm running with a 3090 and 96G RAM, all
619.
▲
by
barrkel
3y ago
The mirror on the wall of the bathroom in the Berlin location looks through to the kitchen in the next room. I guess the depth gauging algorithm uses parallax, and mirrors confuse it, seeming like windows. The kitchen has a blob of blurrine
620.
▲
by
barrkel
3y ago
As I wrote on September 14th, 2001, 3 days after 9/11, "Just don't act rashly". https://groups.google.com/g/borland.public.delphi.non-techni...
621.
▲
by
barrkel
3y ago
25Gbps FTTH is available here in Switzerland, about $75/month (billed annually, 777 CHF). And Switzerland is a country with very high labor prices.
622.
▲
by
barrkel
3y ago
That's not an engineering solution.
623.
▲
by
barrkel
3y ago
If you take a broad view that income is an increase in wealth before expenses, then things are hard to calculate. If you focus narrowly on cash transfers, then you get an unfair tax system, where some forms of wealth accumulation are taxed
624.
▲
by
barrkel
3y ago
That's an easy way to accumulate data corruption. It's better to design for unexpected restarts than design for a golden in-memory image which needs to be carefully ported around, have all its connections wired back up, and so on.
625.
▲
by
barrkel
3y ago
Great, a way to turn videos into something I can scan. Actually something I'd consider using.
626.
▲
by
barrkel
3y ago
Gerrit is a poor man's Critique and it's not nearly as slick to use in practice.
627.
▲
by
barrkel
3y ago
Thanks for policing my snacks. I like to munch on something occasionally of a weekday afternoon while working, or when watching a movie at home in the evening. If it wasn't crisps it would be something very similar: savoury, unheated,
628.
▲
by
barrkel
3y ago
I live in Switzerland nearish (35 minutes) to the German border. I need to drive to France (~1 hour) to get decent flavours. I'm a big fan of Brets. And I'm absolutely sick of Ungarisch, I'll never buy paprika crisps for myse
629.
▲
by
barrkel
3y ago
It's not usually useful, per se, unless you are able to refer to it during programming competitions, or if you're implementing a runtime library and want to brush up on the tradeoffs of some low level algorithms - but even then, t
630.
▲
by
barrkel
3y ago
Thing is, it could only be that way in its early days, when the vanguard of users came to it from word of mouth, from following Joel Spolsky or Coding Horror or their joint podcast. The audience is much bigger now and with the long tail of
More ›