Hacker Newsnew | past | comments | ask | show | jobs | submit | atherton94027's commentslogin

Curious, could you give examples of how you've been able to double your productivity with AI?

In my (not systems engineering) opinion, most time spent writing code is boilerplate and rituals; unit tests are pretty repetitive, creating a React component is a lot of repetition, etc. A LLM code assistant can do these boring things faster.

Yeah I agree with you on that, I'm just curious about the systems programming use case as in my experience you have to think deep about interactions and working with an agent blunts that

What would they be giving up? DoorDash?

By most aspects the world of 1990 didn't change that much from today's world, with the exception of having computers in our pockets and some advances in medicine.


I think you can go back further

I used to work with a guy who would tell me that, except email, life didn’t really change since the 80s. All we did was stick a screen onto everything, whether we needed to or not.

I was only born in the 90s but I mostly agree that far back.


> I think you can go back further

Reminds me of a line by John Maynard Keynes from 1919 about life before WW1 —

“The inhabitant of London could order by telephone, sipping his morning tea in bed, the various products of the whole earth, in such quantity as he might see fit, and reasonably expect their early delivery upon his doorstep”


Are these proprietary drivers owned by American or Asian companies? There are many alternatives to Qualcomm nowadays


86 race conditions compared to what baseline? This is a bit meaningless without benchmarking against other kernels


It's 1 compared to 86, 86 is the baseline.


But you need to control for lines of code at the very least — otherwise you're comparing apples to oranges


I'm perfectly happy to say that it's not a very good way to make a comparison.


Then it would not be unscientific.


Yeah I mean I could also say "there are no CVEs written in PERL in the kernel ergo PERL is safer to write than Rust". Given there's close to zero .pl files in the kernel, I think we can all agree my assertion holds


That claim relies on an absurd "in the kernel" qualifier, making it difficult to agree with. Furthermore, your hypothesis is that "we all" agree with claims that rely on absurd conditions as a matter of course.


That is no base line. That is a comparison with no statistical value.


Tbh I thought that was clear when I used the phrase "unscientific".


Sure this could happen but that seems like a very last resort. The only reason the US economy is still competitive is tech stocks so cutting off ~35% of your income seem like it would cause a lot of downstream effects


This a bit of a hollow article and kind of misses the point of Rob Pike's rant. This is a guy who care deeply about computing. He spent the first 20 years of his career working at Bell labs, building things like utf-8 (an amazing idea that everying easier for anyone who doesn't speak english!) and plan 9. Pike does not like resource wasting (part of the reason they built Go was to replace Python at Google) so yes, his point makes sense in that context


I'm confused, is EBS eventually consistent? I assume that it's strongly consistent as otherwise a lot of other linux things would break

If you're thinking about using NFS, why would you want to distribute your locks across other machines?


Why would anyone want a distributed lock?

Sometimes certain containerized processes need to run according to a schedule, but maintainers also need a way to run them manually without the scheduled processing running or starting concurrently. A shared FS seems like the ”simplest thing that could possibly work” distribution method for locks intended for that purpose, but unfortunately not all cloud storage volumes are strongly consistent, even to the same user, and may take several ms for the lock to take hold.


Wouldn't a database give you better consistency guarantees in that case? NFS locking semantics are a lot more complicated than just a `SELECT .. FOR UPDATE`


Sure, but that would require a separate database for this one use case. Mixing infra concerns into an app db doesn’t sound kosher, either, and a shared volume is already available.

Seems easier to have a managed lockfile for each process, diligently checking that the lock has actually been acquired. Performance is not a concern anyway, as long as acquire takes just a few ms we’re golden.

FWIW, it’s not NFS.


Yeah it honestly feels like the problem here – it's a common pattern where someone tries several times at a promo, then transfers to another team and gets promoted immediately.


Waymo cars have been proven safer than human drivers in California. At the same time, 40k people die each year in the US in car accidents caused by human drivers.

I'm very happy they're moving fast so hopefully fewer people die in the future


Both things can be true. They can be safer, but at the same time Waymo can still externalize stuff to the public...


Who cares? Honestly?


…the public?


There was a special law (that his party voted for btw) that added sentencing enhancements for the specific crime he did iirc


It's surprisingly often the case that criminals assume everyone else is doing what they're doing, and want harsh punishments for it.

(see also: the rates of homosexuality among homophobes)


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: