Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
LtdJorge
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
91.
▲
by
LtdJorge
8mo ago
Adamo never blocks, at least for me. Vodafone does.
92.
▲
by
LtdJorge
8mo ago
Exactly, why does nobody think about this!
93.
▲
by
LtdJorge
8mo ago
I replaced CF tunnels, which kept disconnecting every few minutes with it, and happy.
94.
▲
by
LtdJorge
8mo ago
It’s also tremendously durable, NAND doesn’t get close in write endurance.
95.
▲
by
LtdJorge
8mo ago
Yeah but SRAM is stupidly fast. Even compared to DRAM. Going to a different chip, no matter if over a custom interconnect takes a "lifetime", specially since you don’t have L4 caching tricks, as you’re already at SRAM level.
96.
▲
by
LtdJorge
8mo ago
Hmm, for most desktop stuff, you're still limited to random access, where even if leagues above HDD, the NVMe still suck compared to sequential. It's sad that intel killed Optane/3D X-point, because those are mych better at r
97.
▲
by
LtdJorge
8mo ago
Sounds like typical Windows experience
98.
▲
by
LtdJorge
8mo ago
I didn't want to call it backend, because it's not that. Maybe desktop programmer? Anyway, if you're doing GUI in Rust or similar, you're at least handling memory and syscalls, which is closer to systems programming, I&#
99.
▲
by
LtdJorge
8mo ago
Yes, it's the job of the scheduler
100.
▲
by
LtdJorge
8mo ago
RTP I’d say
101.
▲
by
LtdJorge
8mo ago
And GTK4 is even very usable from Rust too. It’s not a bad development experience, but these companies probably find 100 webdevs for every system programmer.
102.
▲
by
LtdJorge
9mo ago
No, but you have to use a Chromium browser on Windows, otherwise your life will be miserable.
103.
▲
by
LtdJorge
9mo ago
You then use io_uring
104.
▲
by
LtdJorge
9mo ago
Doesn’t ZIP have all the metadata at the end of the file, requiring some seeking still?
105.
▲
by
LtdJorge
9mo ago
Teams inside a VM it is, then.
106.
▲
by
LtdJorge
9mo ago
Use Enterprise
107.
▲
by
LtdJorge
9mo ago
I have an LG OLED C3 as a monitor, 42". I may be able to distinguish separate pixels if looking at a '.' or something like that (a stuck pixel happened for a few weeks, which I could notice on a white background). But the den
108.
▲
by
LtdJorge
9mo ago
If everything is static, they'll cache it in a DC close to you. That's better than what we had before.
109.
▲
by
LtdJorge
9mo ago
Exactly, that’s the point
110.
▲
by
LtdJorge
9mo ago
I know, thats why I used max CPU settings. But when processing map tiles with a final total compressed size of half a terabyte, and each one is 200kB, taking 20s per tiles is prohibitively expensive.
111.
▲
by
LtdJorge
9mo ago
I tried both AOM and rav1e, same quality with rav1e producing 20% larger image, same performance more or less (too long)
112.
▲
by
LtdJorge
9mo ago
If compared to a smartphone, maybe.
113.
▲
by
LtdJorge
9mo ago
AMD doesn't have it. I just confirmed by grepping through dmesg and journalctl -b, the only time it appears is due to UPS driver notifications (unrelated).
114.
▲
by
LtdJorge
9mo ago
To me, the current UK administration, silencing "bad things" and telling you which opinions are right and which are wrong is closer to fascism than the US administration.
115.
▲
by
LtdJorge
9mo ago
I tested both WebP and AVIF with maximum CPU usage/effort. I have not tried the faster settings because I wanted the highest quality for small size, but for similar quality WebP blew AVIF out of the water. I also have both compiled wit
116.
▲
by
LtdJorge
9mo ago
I’ve recently compared WebP and AVIF with the reference encoders (and rav1e for lossy AVIF), and for similar quality, WebP is almost instant while AVIF takes more than 20 seconds (1MP image). JXL is not yet widely supported, so I cannot rea
117.
▲
by
LtdJorge
9mo ago
Do you mean Blackjack?
118.
▲
by
LtdJorge
9mo ago
The Date API is horrible
119.
▲
by
LtdJorge
9mo ago
Yep, I’ve been using this one which is lighter (20kB): https://github.com/fullcalendar/temporal-polyfill/
120.
▲
by
LtdJorge
9mo ago
Yes, but those are done in software
More ›