Krei-se

I was more after the fact i can trust the compiler to give me the same result - even though via an optimized path.

Certainly scopes vary, but in my line of work i define memory layout and how this data will be processed myself - thus it's great a compiler might do that, but the result of the computation will not change.

Now in comparison giving an LLM specs ... i a) cannot be sure what the computation will be b) it might be something else on another run.

I beg to differ because a compiler is deterministic.

PHP is boring - but it works.

Yours a partially smart take for sure.

Once had a coworker adding lots of safeguards to his code - that i knew would never trip - as he was measured by LOC in his prior workplace.

You can even assume 120% is.

These are grown and stable platforms now getting additions with the help of AI (which is not wrong) but they are not built by one.

Its like saying you live in a 3D printed home because you added a 3D printed decoration + the house is structurally stable == its a stable 3D printed house.

also - as always with these claims there's no actual product / repo / whatever one could check.

I would love to see what these tools create but outside slop there's never: This works, is in production, here's the code.

Any day now.

Name them. Just 2! Any example would be enough.

Or well better than 0 and just a claim.

When all else breaks, absurd() is your friend.

So i guess proven secure architecture will come and stay.

all hail the new custom excel spreadsheet, wordpress blog or 1-click shop.

ai coding is the new Symfony w/ bootstrap theme hype cycle. I'll be back to structs. Know your tools!

I like you! Pirsig yes, Zen and the art of motorcycle maintenance - i got from Gemini it's Jackie Stewarts quote though.

Brilliant. You nailed it! It's not the effort you put into it ⸺ it's the (void) thought that counts!

Which borderline suicidal thought-pattern would you like to explore next?

One might think so when alienated from the machine, but you can utilize the right temporal neocortex (mirror of your speech center on the left) which processes and emits emotions for whatever - it does not have to be other humans!

A race driver (dono anymore) called this "Mechanical Sympathy" and it's what happens when you work on what feels wrong / tedious etc when working on your computer. And that usually involves doing things down to the bits.

Can you look into your case on a memory bank and tell what chips hold what data? How it flows from the input of your fingers to the screen? A small C program using the framebuffer is within reach!

It's lovely!

It's because of all the added abstractions. 20-30 years ago you were closer to the metal, similarly to how you would play with arduino now. But now all you get is blackboxes which fogs the mind.

De-alienate yourself! Feel the tools again! Learn low-level concepts and use them in your work (data-driven design, functional programming, etc.).

Then it's like woodworking and cooking again.

Hit me up if you want to build a 2000-era (V)LAN:

- XMPP Chat

- Layer 2 VPN (IPX!!)

Then let's game!

You can use xkb with just 5 concepts thus reducing the complexity from ~20 to 5:

- scancodes (evdev) -> keycodes

- keycodes --(keytype)--> keysymbol (actual character).

- modifiers actions -> keytypes

Useful but not needed: Keycode-aliases, indicators

Scrapped: All .lst, .xml, Rules, Models, Layouts, Variants, Options, virtual modifiers, explicit, implicit, indirect.

It's possible to write a full X11 compatible map with 8 modifiers, 4 groups, 8 levels getting (108) * 4 * 8 = 3456 characters. Then you only send this map and actually know whats going on - never needing strange overrides, etc. - which feels great.

It took me 4 days, you can likely do it in 1-2.

Thanks for sharing that experience, i often wondered whether I'm alone being lead on ad infinitum.

BUT - i wrote a small tool recently where scope and environment was set and i wrote the code myself so just wanted a security check. I followed through with all suggestions just to see if it may end - ...

it does!! After adding XSS, CSRF and fixing all undefined edge cases it was actually happy!

So i'd say yes if you define the scope clearly then it will get you there.

Your host with the webserver has no idea its just a docker image. It talks to app.domain.tld and gets a reply. app.domain.tld asks db.domain.tld with an SQL query and gets a reply. But it can be one of any deployed docker image on any bare metal host - which one is db. and app. etc. is decided by Kubernetes.

In the background kubernetes routes all these docker images with each other without you having to think about this. You may have 1000 db.domain.tld nodes caching a master db host - if you configure that in the docker image, that's not different to bare metal replication.

Same with load balancing in webservers. You can do that! Or you can just have kubernetes handle it. I'm not sure but would expect it to swap images in a way network is optimized - i don't use this kind of software. I just know it's done like this because the bottleneck of modern software is not the local network so its not an issue to have these pieces on different bare metal hosts. And its easier to stay operative if some hosts fail - but you can all solve this by hand.

I work with bottlenecks between network, ssd, ram and vram but if you deploy npm riddled software for >1 Million users well then you may want Kubernetes. Or if you are google and have 10k engineers that need to agree on a standard!

If you can do this yourself with load balancing, replication etc - do it yourself and don't think there's something wrong with that. It's more elegant, efficient - but you have to agree with others how you do it. And that can also be a bottleneck ;)

Still i think you doing it by hand is better so don't worry you are not old you may just have higher standards.

Oh im sorry i had no idea to talk to other hackers i have to go to usernews.com obv. when this is the wrong place to think like a techie.

If you achieve 100% delivery rate you will surely share the results, right? No? Oh i guess all those services simply tell you they make it easier for you while not actually achieving that - but why check?

Same. People just remember it's hard because they don't test, don't like responsibility much and esp. because they test this in production not on a spare domain. Then all you remember from it is the 10 people shouting at you and that you need to fix it NOW and it's all super stressful so they never attempt this again.

Have you actually measured delivery rate of some service vs. doing it yourself?

"Hacker"news: Don't actually do it yourself man, just pay someone!

you can check your delivery rate with services there are lots that are free to check and they also offer to "warm" your reputation. I have not done that yet (money!) and outside this:

- Outlook.com consumer mail addresses

you can still get >99% delivery rate if you have spf, dkim and all the new cool stuff.

So no, please DO NOT recommend against doing it because ONLY if mail stays a decentralized service providers will stay forced to not blacklist everything but other large providers.

I self-host with telekom business, 2 fixed ip lines fiber and vdsl at home everything from auth ns to mail and web. Its around 85€ for me.

it's still doable and you get the best feeling there is: Living in the internet.

Can recommend 100%.

i host myself with 2 fixed IPs from telekom germany business, its not as expensive as one might expect, under 90€ for a fiber and vdsl line, so next to hosting your own auth ns servers and mail you get failover at home which is nice.

Yeah you're not wrong, that wasn't really necessary and is no way to treat people that may make good engineers.

See if you use your time to write specifications so AI will stay on guardrails - why not just go all the way to memory layout and functional programming? It's great to know what actually happens on the machine.

With natural language or the 4323th framework that makes things "easier" you just end up yourself not actually knowing what your intent is your machine should fulfill. Shouting at some AI the stuff your boss shouts at you? Not exact logic and what people intend today may contradict what they wanted yesterday.

Having domain knowledge and knowing how to translate peoples immediate wishes, predicting what they will want after that into building solid software is a skill you can grow into, but switching the toolset all the time that hides the magic keeps you from learning what was true in the 70s and is true today.

No one has to agree with this just because torvalds said it.

You can also read up here: https://lore.kernel.org/linux-media/20260715125058.3692aa03@... that this thread is about code analysis not generating it and the signal/noise ratio. It's not like his opinion is shared by everyone because these people actually have to look at the PRs generated and filter out all the noise.

The issue is not AI finding bugs is bad, it's that usually a maintainer gets offloaded 1000s of automatically generated PRs where 99,9% are BS and is then expected to read and check and understand all of them, immediately merging the 0,1 actually pointing out a real security issue.

You working in some "we replaced MS Access with another high-level language tool"-place is so far off this level of engineering i'm not surprised you only took the information "AI good" from this then told everyone else about how wrong they are to not use it.

You're not wrong but there's no need to digress further than the project he manages. Also many seem to think it's wrong to like some parts about a public figure (which he clearly is) while not agreeing with everything he says. You don't have to like his stance on this!!

He's not writing code, he merges 200 PR in 2 weeks. From this POV an AI eases his workload, but he makes no difference whether he dislikes AI generated code or is just fine with found bugs by it.

Still he has no idea whether the person suggesting a feature or fix has any actual knowledge about the code so when problems come up there's no one there taking responsibility and i think that's shortsighted.