Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
zokier
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
16 ms
·
271.
▲
by
zokier
1y ago
What is your complaint here exactly?
272.
▲
by
zokier
1y ago
Title is fine. The key is the plural, smartwatche s . From the actual article: > No geodetic station is involved in this study, as one of the smartwatches serves as the base station.
273.
▲
by
zokier
1y ago
Well they also claim to be able to cache build steps somehow build-system independently. > As the build runs, any step that exactly matches a prior record is skipped and the results are automatically reused > SourceFS delivers the per
274.
▲
by
zokier
1y ago
There are like at least half dozen open source terminal emulators that have decent performance text rendering if you want to look into how it is really done. It is not simple, but at this point I feel it is a largely solved problem.
275.
▲
by
zokier
1y ago
> But I'm unaware of rendering engines that do alpha blending in something other than linear or SRGB. Well, spectral rendering is a thing, kinda bypasses the problem of color blending for rendering in some cases.
276.
▲
by
zokier
1y ago
Kingbright recently released 01005 (0.45mm x 0.25mm x 0.2mm) sized LEDs, which afaik are one of the smallest ones easily available. One neat idea would be to pack those on DIP14 sized pcb, making tiny neat character display. I guess somethi
277.
▲
by
zokier
1y ago
Those look like normal 2.5-3mm LEDs, that is big difference to 1mm² LEDs. The circle disc from OPs link has 2.5x higher LED density, and they could be probably packed even more densely in a grid.
278.
▲
by
zokier
1y ago
Elastic II is very neat by preserving drainage basins. But Antarctica looks really distorted, in a way that doesn't seem necessary? Could that be fixed somehow?
279.
▲
by
zokier
1y ago
for blending oklab almost always works better than srgb (linear or gamma).
280.
▲
by
zokier
1y ago
I'd assume it is 64-bit, which would explain why it is limited to Pi 3 upwards
281.
▲
by
zokier
1y ago
Doesn't nvme have lot of log capabilities, like telemetry log etc?
282.
▲
by
zokier
1y ago
Personally I think these problems are a sign that posix fs/io apis are just not that good. Or rather they have been stretched and extended way past their original intent/design. Stuff like zenfs give interesting glimpse of what co
283.
▲
by
zokier
1y ago
The difference is that Windows 11 was released 4 years ago, and Windows 10 just reached EOL. And we havent heard anything yet about Windows 12, and MS still seems very intent on doubling down on their chosen path. In contrast Windows 7 was
284.
▲
by
zokier
1y ago
Windows 11 has significantly worse adoption rates than 10 or 7, more comparable to the notoriously poor Vista. It is not just few HNers complaining here. https://arstechnica.com/gadgets/2023/10/windows-11s-ado
285.
▲
by
zokier
1y ago
I have been using forgejo for a very small sideproject of mine and so far I have enjoyed it, but admittedly I use probably like 10% of the features. But sometimes that is all you need, and forgejo is quite nice and simple to configure and
286.
▲
by
zokier
1y ago
straight from the abstract: > we can demonstrate quantum-amplified time-reversal spectroscopy on an optical clock transition that achieves directly measured 2.4(7) dB metrological gain and 4.0(8) dB improvement in laser noise sensitivity
287.
▲
by
zokier
1y ago
Relative to the effort that has been poured to volte etc and 3g/4g/5g in general.
288.
▲
by
zokier
1y ago
Penny pinching. Afaik asymmetric PON is the cheapest possible network tech at scale.
289.
▲
by
zokier
1y ago
Aren't most residential fiber deployments PONs which generally do not offer symmetric bandwidth? E.g. 10G-PON has 10G down / 2.5G up.
290.
▲
by
zokier
1y ago
In lot of contexts scaling down is far more important than scaling up. In that sense scalability is cost-optimization; instead of provisioning fixed capacity that is enough for (predicted) peak loads, you can scale based on actual demand an
291.
▲
by
zokier
1y ago
Incoming calls should be subject to acls and default-deny policy should be practical. This means that - caller identity should not be spoofable - identities should form hierarchies and groups so you can allow whole organizations instead of
292.
▲
by
zokier
1y ago
I feel the fundamental problem is the idea of phone number being secret or private in the first place. It is just very weak form of access control, I believe we can do much better these days.
293.
▲
by
zokier
1y ago
> can't be inspected by arbitrary functions in the same process You can not really rely on that. The initial stack layout is well-defined at least on linux, so digging up argv is not difficult. Or just open /proc/self/
294.
▲
by
zokier
1y ago
5G has fairly sophisticated positioning support, especially in the latest releases. Base stations generally are in precisely known locations and have good timebases. Urban locations which are challenging for gnss usually have good 5G covera
295.
▲
by
zokier
1y ago
Isn't the software (licensing) situation for second-hand enterprise switches bit iffy?
296.
▲
by
zokier
1y ago
It is remarkable that this is accomplished with relatively modest setup and effort, and the results are already great. Makes me wonder what you could get with high-end gear (e.g. 61mp sony a7rv and the new 100mm 1.4x macro) and capturing mo
297.
▲
by
zokier
1y ago
Not necessarily sqlite, but rocksdb is looking like promising option for foundational storage system. You have things like zenfs which runs rocksdb directly on nvme storage, and then you have lot of higher level systems on top of it like Ce
298.
▲
by
zokier
1y ago
In theory it allows to pack text tighter with minimal line spacing.
299.
▲
by
zokier
1y ago
AWS has this "mountpoint for s3" thingy https://github.com/awslabs/mountpoint-s3
300.
▲
by
zokier
1y ago
But how much is a 25GbE (or 40GbE) switch?
More ›