HN user

fzeindl

1,139 karma

Fabian Zeindl

Senior Software Architect @ Austrian railways ÖBB

https://www.fabianzeindl.com/

-- Blog

* The API database architecture - Stop writing HTTP-GET endpoints - https://www.fabianzeindl.com/posts/the-api-database-architecture

* The business information server: generic data checking using PostgreSQL, PostgREST, and Semgrep - https://www.fabianzeindl.com/posts/business-information-server

* The code quality pyramid - https://www.fabianzeindl.com/posts/the-codequality-pyramid

* ChatGPT is able to simulate age groups - https://www.fabianzeindl.com/posts/chatgpt-simulating-agegroups

* Technical debt and technical clutter - https://www.fabianzeindl.com/posts/technical-debt-and-technical-clutter

Posts20
Comments244
View on HN
www.eff.org 21d ago

Supreme Court says constitution protects people's location data

fzeindl
6pts0
bottleneck.jakubgornicki.com 2mo ago

Bottle Neck: Strait of Hormuz Crisis Simulator

fzeindl
2pts0
www.dw.com 3mo ago

Austria sends 70k kids on a three‑week smartphone detox [video]

fzeindl
3pts0
www.derstandard.at 4mo ago

72.000 pupils in Austria start 3-week smartphone-detox in controlled study

fzeindl
1pts4
infrastruktur.oebb.at 7mo ago

Koralm Railway

fzeindl
317pts197
news.ycombinator.com 11mo ago

Ask HN: What is wrong with modern software development

fzeindl
6pts9
news.ycombinator.com 11mo ago

Ask HN: Why don't LLMs replace bosses instead of engineers?

fzeindl
18pts19
github.com 1y ago

Kotlin JSON Stream Serializer

fzeindl
2pts0
www.fabianzeindl.com 1y ago

Type-safe, multiplatform JSON serialization for Kotlin generated at compile-time

fzeindl
3pts0
news.ycombinator.com 1y ago

Ask HN: Does anyone use unattended security upgrades inside Kubernetes-Pods?

fzeindl
1pts0
www.bestpractices.dev 2y ago

OpenSSF Best Practices Badge Program

fzeindl
1pts0
www.fabianzeindl.com 2y ago

The API database architecture – Stop writing HTTP-GET endpoints

fzeindl
93pts164
ai.stackexchange.com 2y ago

How can a LLM optimise on it's own processing speed?

fzeindl
1pts0
mcmansionhell.com 3y ago

McMansion Hell

fzeindl
86pts78
spaceexplored.com 3y ago

Scientists find phosphorus in an ocean of Saturn’s moon Enceladus

fzeindl
4pts0
xosh.org 3y ago

Online Text to Diagram Tools

fzeindl
2pts0
www.fabianzeindl.com 3y ago

The Code Quality Pyramid

fzeindl
1pts0
www.fabianzeindl.com 3y ago

Generic data checking with PostgreSQL, PostgREST and semgrep

fzeindl
2pts0
www.fabianzeindl.com 3y ago

ChatGPT is able to simulate age groups

fzeindl
2pts0
arstechnica.com 3y ago

How fighting games use delay-based and rollback netcode (2019)

fzeindl
121pts92

Isn‘t this an example for something that will pay off later? Public baths come to mind. Those generally don‘t make sense from an economic point of view and are prohibitively expensive for private owners to build and maintain.

But they pay off by keeping society clean and healthy and prevent loss of workforce from mental or physical disease?

In many cases the skills are available in house to do the necessary vetting, but these people are already overwhelmed with their existing day to day.

This is an interesting topic. We treat vetting output the same as doing the work ourselves, but that is not the case.

Doing the work is not the same as reviewing work done by others.

I have heard reports of software engineering companies that have gone full agentic. Their seniors only review stuff written by LLMs and it burns them out, because they have to switch context constantly.

I find this interesting because part of being a senior developer is that you are experienced enough that you won‘t make grave mistakes anymore. This is the case in many professions: you are relied upon to not make grave mistakes.

But those same people are now swamped with stuff that they are not able to review, so they will let a grave mistake slip through at some point.

So they really can‘t trust themselves anymore?

but I am absolutely in the camp of "there are many tasks where pattern matching is just as effective as actual understanding“

What if „being effective at something with pattern matching but not understanding it“ just means that you have identified only 90% of patterns and keep failing to learn the rest for whatever reason.

I was originally sceptical of LLMs and am far from the „agents will magically fix our future“-crowd, but sentences like these trip me up:

„But pattern‑matching is not system understanding, and plausibility is not correctness.“

Why not? Who says that? Who proved that system understanding is not just more complex pattern matching?

„LLMs predict tokens, not consequences“

Same here. LLMs output tokens but who says that they don’t form some internal group of token-predicting tensors that move together and constitute the internal model of a „consequence“? It is like saying humans don’t have thoughts, they just have electrical impulses moving their tongues.

I too think that LLMs seem to be a very specific form of intelligence, maybe resembling the parts of our brain that do language-processing, but it is a fact that they at least fake intelligence very convincingly. And that we actually don’t know how they do it.

It's true that there are few obvious bugs, but there are many subtle bugs now, usually outside the main interaction flows. 5 years ago, apple's software felt bugfree, essentially.

examples:

- if I change a note on my iphone and wake up my mac, I need to restart the notes-app before it syncs the change.

- if somebody leaves me a FaceTime video-message, I get an "unread"-badge that doesn't get away after I watch the video. There are multiple ways to get to that video and only one of them clears the "unread" badge.

- if I add a pronounciation field to a contact in my iphone, SIRI stops working and I need to restart my iphone to get it back.

I exercise 3 times a week (running, martial arts, weight lifting). I ran marathons, I exercised 5-6 times per week for the last 5 years.

I am truly fit, but … I still don’t feel like exercising.

Most of the time it’s like brushing my teeth, it is just something I do because it is right.

No.

With databases there exists a clear boundary, the query planner, which accepts well defined input: the SQL-grammar that separates data (fields, literals) from control (keywords).

There is no such boundary within an LLM.

There might even be, since LLMs seem to form adhoc-programs, but we have no way of proving or seeing it.

The principal security problem of LLMs is that there is no architectural boundary between data and control paths.

But this combination of data and control into a single, flexible data stream is also the defining strength of a LLM, so it can’t be taken away without also taking away the benefits.

But a lot of UK housing relies on on-street parking, and there's flats with car parks where charging isn't currently practical.

You forget the larger problem less wealthy individuals face: They typically already own a ICE-car and can‘t afford to purchase a new car multiple times in their lives.

Day 1 of ARC-AGI-3 4 months ago

Would the single sentence „Imagine you are a regular computer player and accustomed to the usual elements of games“ count as a harness?

a lot of real world performance comes down to how efficiently you are calling external services (including the database)

Apart from that my experience over the last 20 years was that a lot of performance is lost because of memory allocation (in GCed languages like Java or JavaScript). Removing allocation in hot loops really goes a long way and leads to 10 or 100 fold runtime improvements.

Java 26 is here 4 months ago

I agree JFR is amazing.

Also that you can start it on a running JVM and it that is has minimal ~1% performance overhead.

Java 26 is here 4 months ago

That and also the JVM is a marvel of engineering with millions of hours that were poured in by smart people over the last 30 years.

It can be as fast a C, supports a multitude of languages, introspection, surveillance etc.

The MacBook Neo 4 months ago

My guess is that some cell in an excel sheet says that some customers bought certain models in the past and no manager at Dell has enough weight or enough courage to question that and rule to NOT release a certain model.

I have been alone for quite some time in my life, for various reasons.

Here is my advice

1. Don’t listen to people telling you that there‘s something wrong with you if you aren’t happy alone and that you need therapy. You probably don‘t.

If you aren’t able to survive a single day on your own, then you have an issue, but in general people are social animals and need talking to and touching other people on a regular basis to regulate themselves emotionally. Some don’t, but you don‘t seem to be one of them. (In my experience people who honestly enjoy being alone for weeks at end often got used to this in their childhood, for example by being single children. If you had a noisy, lively house when growing up you will always miss that.)

2. That being said it is still nourishing and useful to follow the tips for spending time on your own (cooking for yourself, going into the cinema) now and then. But don‘t expect it to magically fix your yearning for social connections.

3. Doing sports and going to the gym often is great advice, because sports generally gives you a more positive mood and a better outlook and it quiets down negative thoughts. But also keep in mind here that it will not replace social interactions and touch.

4. So the question remains where you get your social interaction from, and I suggest that you replace the single person you had with a lot of tiny social interactions.

None of those is going to fix your solitude on it’s own, but the mixture will enrich your life. If you do all of those things once per week, your mood will improve:

- Online dating and chatting with friends (but don’t expect anything to come out of it)

- Dancing classes (this is especially great because you get the physical touch component)

- Regular coffee/dinner dates (1-2 per week) with a friend, coworker or relative.

- Going for a walk and talking to a random stranger

- Singing in a choir or similar hobbies

- Going to pubquizzes or free to join community walks

- Commuting to the office instead of doing home office.

5. I won‘t argue against your psychiatrist, but I find that medication is only useful in a minority of cases. Both anti depressants and anti anxiety meds can actually reduce your motivation to go out, by mellowing you into a not-great-not-terrible dreamy state.

6. Don‘t start rummaging and contemplating your thoughts when you are lonely. Being lonely makes you weird: https://arstechnica.com/science/2023/07/lonely-people-see-th...

I once digged into this database out of curiosity and found incredibly detailed research on many edge cases. Like time zones in Germany being temporarily aligned to Moscow during soviet occupancy after World War Two.

One particular commenter stood out to me, so I looked him up because I was interested which kind of people spend so much time correcting timezone information.

Turns out he was an astrologer and wanted his astrology-program to work perfectly correct.

I find it funny that we have to thank astrology for the correct calculations in our banking software :).

Almost no one is shopping multiple supermarkets, buying cheaper alternatives, buying in-season veggies and fruit when it’s cheap, waiting for sales to stock up, buying in bulk and freezing, using coupons, meal planning based on the latest supermarket Sunday circular, etc. only a tiny minority of people have been doing so.

I don‘t know where this observation comes from, but here in Austria a majority of people in lower income sectors than IT do all of this?

In the book „the design of everyday things“ it is mentioned that „natural mappings“ exist. Moving the knob of a vertical slider to the upper end universally means „brighter“ or „louder“, not „less bright“ or „more silent“.