Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
nunez
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
33 ms
·
751.
▲
by
nunez
10mo ago
Porn is a category; apps are a concept Like age laws for vape pens vs age laws for shopping.
752.
▲
by
nunez
10mo ago
Between AI improvements, laws like this and Telly, we are a few steps away from the telescreen. (I saw a Telly recently. This device should be terrifying, but "free" makes people make weird choices.)
753.
▲
by
nunez
10mo ago
Yes, but you can't stop eight year olds from grabbing a James Patterson or Stephen King novel from the shelf. Their parents should, and some librarians might throw a moral exception to their choice, but if they wanna read It, they'
754.
▲
by
nunez
10mo ago
The file permission system on Windows allows for super granular permissions, yes; administrating those permissions was a massive pain, especially on Windows file servers.
755.
▲
by
nunez
10mo ago
Correct. I was shocked when one of my non-technical family members moved over to Proton Mail. I was super proud of them even if it came from left field!
756.
▲
by
nunez
10mo ago
Used this all of the time back in the day. Great tool.
757.
▲
by
nunez
11mo ago
lol all of the latter points are true though (except for it being only men, though i get how things were back in the day)
758.
▲
by
nunez
11mo ago
Yeah, but being able to produce nuclear-sized 10k+ LOC PRs to open-source projects in minutes with relatively-zero effort definitely is. At least you had to use your brain to know which blog posts/SO answers to copypasta from.
759.
▲
by
nunez
11mo ago
> 50k line PRs that required a 10-20 line change This is hilarious. Not when you're the reviewer, of course, but as a bystander, this is expert-level enterprise-grade trolling.
760.
▲
by
nunez
11mo ago
I feel like a story about some open-source project getting (and rejecting) mammoth-sized PRs hits HN every week!
761.
▲
by
nunez
11mo ago
This sounds like what unit tests after every commit and e2e tests before every PR are supposed to solve.
762.
▲
by
nunez
11mo ago
Red Hat launched an equivalent effort in October: https://www.redhat.com/en/technologies/linux-platforms/enter...
763.
▲
by
nunez
11mo ago
Yes IMO
764.
▲
by
nunez
11mo ago
It becomes a problem if you're testing something in local Kubernetes clusters that are ephemeral
765.
▲
by
nunez
11mo ago
Case in point: subscription costs for everything going up to justify the "additional value" that AI is bringing _whether you use it or not_. This would have been an additional, more expensive, subscription tier in the past. Anecdo
766.
▲
by
nunez
11mo ago
> However tech people who thinks AI is bad, or not inevitable is really hard to understand. I disagree. It's really not. Popular AI is extremely powerful and capable of a lot of things. It's also being used for nefarious purpos
767.
▲
by
nunez
11mo ago
Yes, and I do that aggressively. But that's an opt-in behavior, and like most things that are opt-in, people can't be bothered, and companies know that.
768.
▲
by
nunez
11mo ago
I sympathize with this so deeply: > "Requiring a smartphone to exist in society is not inevitable." Seeing smartphones morph from super neat computer/camera/music players in our pockets to unfiltered digital nicotine
769.
▲
by
nunez
11mo ago
It's an easy process; I wrote up how to do it: https://gist.github.com/carlosonunez/a0ec3f02576867329bc313b...
770.
▲
by
nunez
11mo ago
The interior is more or less the same as the ICE version of the F-150, and there is a lot of parts sharing between them. Probably part of the reason why it didn't have great range. EVs are also much simpler to repair, on average. Defin
771.
▲
by
nunez
11mo ago
Wow; that's a 101-level exploit.
772.
▲
by
nunez
11mo ago
This paradox almost doesn't matter. Non-technical business stakeholders who own requirements for line-of-business apps can generate *working* (to them) end-to-end prototypes just by typing good-enough English into a text box. Not just
773.
▲
by
nunez
11mo ago
LLMs (well, most of the frontier and popular open-source models) are actually quite good at abiding by weird formats like this given that your prompt describes them clearly enough. The real problem is that you'll have to manually spot-
774.
▲
by
nunez
11mo ago
Who asked for this? I thought Amazon was all about customer obsession, and I'm having a hard time imagining readers saying "You know? This book would SEND ME if it had a chat assistant."
775.
▲
by
nunez
11mo ago
You hit this real quick when trying to build container images from the scratch. Theoretically you can drop a Go binary into a blank rootfs and it will run. This works most of the time, but anything that depends on Go's Postgres client
776.
▲
by
nunez
11mo ago
I think they made some splashy hires at the time, and they contributed to the Google SRE workbook. Same as Walmart. They definitely tried. Corporate inertia is a killer.
777.
▲
by
nunez
11mo ago
This is sorted mostly alphabetically with an allowance for people being bad with fractions. That's my guess.
778.
▲
by
nunez
11mo ago
GitHub does! They tell you when you pushed something dangerous almost right away. GitHub Advanced Security blocks the push, I believe.
779.
▲
by
nunez
11mo ago
Tesla FSD has been more or less the same experience. Some people drive 100s of miles without disengaging while others pull the plug within half a mile from their house. A lot of it depends on what the customer is willing to tolerate.
780.
▲
by
nunez
11mo ago
No. The Watch will gladly unlock your other stuff if it's unlocked regardless of the wrist it's on.
More ›