Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
jatone
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
18 ms
·
181.
▲
by
jatone
5y ago
unlikely. electricity is easier to control safely than gas. you don't need to start out full blast with electric cables. ramp up over a minute if necessary. we are also well versed in circuit breakers, and digital safety checks. oh the
182.
▲
by
jatone
5y ago
depends on a lot of things, including how the battery charging is laid out in the battery itself. if you have 100 cells and can charge them simultaneously vs 100 cells where you have to charge sequentially you're going to see huge diff
183.
▲
by
jatone
5y ago
this was always possible.... ssh just didn't do it. but you absolutely could use your key to sign data.
184.
▲
by
jatone
5y ago
> attorneys do not typically need to revert and once a document is executed, prior drafts aren’t really that important. this isn't 100% true. lawyers very much care about changes in legal codes over time. its incredibly time consumi
185.
▲
by
jatone
5y ago
laughs in dvorak
186.
▲
by
jatone
5y ago
my favorite is often being the only developer on linux and giving two files with different casing and watching their systems crash and burn.
187.
▲
by
jatone
5y ago
divorce court absolutely can make the determination that there was domestic violence in their ruling. it simply doesn't have criminal implications. there is very frequently different outcome in civil suits vs criminal suits. where you&
188.
▲
by
jatone
5y ago
just use pkg/errors? it includes the stack trace. there wasn't a need to include this in the stdlib thats all.
189.
▲
by
jatone
5y ago
ive been trying to get them to fix vendoring and modules. they've been incredibly obtuse about the whole situation. =/
190.
▲
by
jatone
5y ago
its really not as bad as people say. linux has same pretty reasonable standards that DEs all implement. most things you need for a DE run as daemons. the biggest pieces that cause problems are UI theming and application distribution. themin
191.
▲
by
jatone
5y ago
yes, and they are absolute nightmares to work with. which is fine its their project.
192.
▲
by
jatone
5y ago
eh you don't seem to realize how shitty current (GTK/Qt) toolkits are from developer ergonomics.
193.
▲
by
jatone
5y ago
its really not that hard. existing stuff make this a fairly easily solvable problem. all you need for a DE is basically the wayland server, an application launcher, system tray, and settings programs. I've been building one myself over
194.
▲
by
jatone
5y ago
jesus people they said they hope it stays: aka socially it becomes normalized. No one is advocating forcing people to wear masks forever.
195.
▲
by
jatone
5y ago
rolling updates isn't actually a problem in practice. the maintainers do test the applications before releasing the updates. i encounter usually 2-3 bugs (and almost always they are minor) per year due to rolling updates and usually it
196.
▲
by
jatone
5y ago
this problem is precisely what tooling is for. golang for example solves the explicit import problem quite nicely with almost no friction outside of ambiguous cases, which you'd need to handle explicitly anyways.
197.
▲
by
jatone
5y ago
they were very up front about it.... 'operation light speed' they were explicitly removing barriers to speed up development and deployment.
198.
▲
by
jatone
5y ago
thats news to me.... you sure some extension you installed isn't adding them?
199.
▲
by
jatone
5y ago
fun fact: you already are in linux. being a monolith doesn't change the nature of the problem.
200.
▲
by
jatone
5y ago
they're not wrong. its not the software that's the issue, its the data stored on disk. the transaction ID is stored in the data for various reasons. in theory nowadays it wouldn't be too hard to change if you use logical repl
201.
▲
by
jatone
5y ago
modern software spans the globe, often communicating with 3rd party systems, operates 24/7/365, and is often at the whims of deadlines. its some of the most technically challenging endeavors in existence. its not really surprising
202.
▲
by
jatone
5y ago
jesus my updates take < 30 seconds on linux though I update frequently. but even on my systems where I don't upgrade for over a years they only take about 10-15minutes.
203.
▲
by
jatone
5y ago
eh having used both, rust's approach isn't any better. when you need to decorate an error, which generally is a lot of the time, rust has the same ergonomics as golang. they optimized code for a generally rare situation. I do appr
204.
▲
by
jatone
5y ago
this is simply because a single implemention doesn't and won't meet all needs. many times you need a simple component that just renders properly, but some people need a11n, others need a bunch of i18n which opens up another can of
205.
▲
by
jatone
5y ago
lack of support for nvidia is entire nvidia's own fault. basically they decided to go there own way on buffer management for userspace and the development community rejected them at large. it'll be fixed in about 1 year, nvidia fi
206.
▲
by
jatone
5y ago
huh? how would a random file be more trust worthy than metadata inserted by the build tool itself at build time? only way for this to be 'incorrect' would be if someone intentionally fucked it up or disabled it.
207.
▲
by
jatone
5y ago
no its not. build tools can use whatever names they want for environment variables. sharing tends to cause more issues then having namespaced ones due to unintended side effects.
208.
▲
by
jatone
5y ago
I mean honestly who cares? its their time. I use rust over c/c++ these days for practical reasons. 1. the toolchain is better. (cargo ala) 2. maintenance of the software is better. (static linked binaries, easier deploys) 3. memory saf
209.
▲
by
jatone
5y ago
thankfully go generics seem far better thought out.
210.
▲
by
jatone
5y ago
wait until you learn about signed bytes.
More ›