Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I’m spending months coding the old way

miguelconner.substack.com

Post reply on HN
Print view View original

151–160 of 387 posts

Re: I’m spending months coding the old way

#151

Post by inerte »

fouronnes3 wrote: post #149

I wonder if we could design a programming language specifically for teaching CS, and have a way to hard-exclude it from all LLM output. Kinda like anti virus software has special strings that are not viruses but trigger detections for testing. This would probably require cooperation during model training, but now that I think of it, is there adversarial research on LLM? Can you design text data specifically to mess w…

> Can you design text data specifically to mess with LLM training?

Maybe text that costs a LOT of tokens. Very, very verbose. I think if there are rules and on the internet, LLMs can eventually figure it out, so you have to make it expensive.

Another way would be to go offline. Never write it down, only talk about it at least 50 meters away from your phone. Transmitted through memory and whisper.

Re: I’m spending months coding the old way

#152

Post by scarface_74 »

Terr_ wrote: post #141
scarface_74 wrote: post #126

Earlier quoted context omitted.

My responsibility is to make sure my code meets functional and non functional requirements. It’s to understand the *behavior*. My automated unit, integration, and load tests confirm that. Someone thought I was naive when I said my vibe coded internal web admin site met the security requirements without looking at a line of code. I knew that because the requirements were that anyone who had access to the site could do…

As written, I do think that's naive. Being sure the person/browser is authorized doesn't mean that the signals you get are actions they intended. Suppose that in normal use a user can visit a certain URL which triggers a dangerous effect. An attacker could trick the user into performing the action by presenting a link to them titled "click here for free stuff." There are various ways to protect against that (e.g. COR…

And that same user is already trusted to have admin access to the entire organizational AWS credentials - I did say it was an internal management site.

The lambda itself only has limited permissions to the backend. The user can’t do anything if the lambda only has permission to one database and certain rights to those tables, one S3 bucket, etc.

Heck with Postgres on AWS you can even restrict a Cognito user to only have access to rows based on the logged in user.

And the database user it’s using only has the minimum access to just do certain permissions.

Re: I’m spending months coding the old way

#153

Post by scarface_74 »

losvedir wrote: post #146
scarface_74 wrote: post #126

Earlier quoted context omitted.

My responsibility is to make sure my code meets functional and non functional requirements. It’s to understand the *behavior*. My automated unit, integration, and load tests confirm that. Someone thought I was naive when I said my vibe coded internal web admin site met the security requirements without looking at a line of code. I knew that because the requirements were that anyone who had access to the site could do…

It wouldn't prevent the admin page from exfiltrating data, though, right? Like, POSTing whatever data is loaded on the page to an arbitrary attacker controlled website.

That would require the logged in user to do something stupid. That’s like saying what’s to prevent the authorized user from emailing his credentials to a random person.

Re: I’m spending months coding the old way

#154

Post by bsder »

JuniperMesos wrote: post #81
LeCompteSftware wrote: post #21

This is ominous and very depressing given what we've recently learned / reconfirmed about LLMs sapping our ability to persist through difficult problems: > There were 2 or 3 bugs that stumped me, and after 20 min or so of debugging I asked Claude for some advice. But most of the debugging was by hand! Twenty whole minutes. Us old-timers (I am 39) are chortling. I am not trying to knock the author specifically. But he…

Why shouldn't someone consult some kind of external resource for help, after struggling with a specific coding problem for 20 minutes? Why is 6 hours the right amount of time to timebox this to?

1) 20 minutes is barely enough time to get into flow.

2) There are different levels of debugging. Are your eyes going to glaze over searching volumes of logs for the needle in a haystack with awk/grep/find? Fire up the LLM immediately; don't wait at all. Do the fixes seem to just be bouncing the bugs around your codebase? There is probably a conceptual fault and you should be thinking and talking to other people rather than an AI.

3) Debugging requires you to do a brain inload of a model of what you are trying to fix and then correct that model gradually with experiments until you isolate the bug. That takes time, discipline and practice. If you never practice, you won't be able to fix the problem when the LLM can't.

4) The LLM will often give you a very, very suboptimal solution when a really good one is right around the corner. However, you have to have the technical knowledge to identify that what the LLM handed you was suboptimal AND know the right magic technical words to push it down the right path. "Bad AI. No biscuit." on every response is NOT enough to make an LLM correct itself properly; it will always try to "correct" itself even if it makes things worse.

Re: I’m spending months coding the old way

#155

Post by scarface_74 »

grebc wrote: post #138
raw_anon_1111 wrote: post #110

Earlier quoted context omitted.

So the project I mentioned while I did write every single line of app code and IAC, made every architectural decision, etc., I did come on an off the project over the course of a year and I couldn’t even remember some of the decisions I made. I also used Codex and asked questions about how the codebase worked to refresh my own memory. Why wouldn’t a junior developer do the same? I mentioned that I had Codex describe…

If you’re 50+ as you intimated in your first post then you have a wealth of knowledge that juniors don’t. Just because it worked this time doesn’t mean it always will. If you need further explanation of why you might want to spend more time resolving a bug to learn about the systems you’re tasked with maintaining then I’m at a loss sorry.

And then as experience developer you would have to try one of the other tools in your toolbox. Why should someone tie a hand behind their back and not use an LLM out of some sense of nerd pride?

Re: I’m spending months coding the old way

#156

Post by ChrisMarshallNY »

apricot wrote: post #86

I am this very term teaching 18-year-old students 6502 assembly programming using an emulated Apple II Plus. They've had intro to Python, data structures, and OO programming courses using a modern programming environment. Now, they are programming a chip from the seventies using an editor/assembler that was written in 1983 and has a line editor, not a full-screen one. We had a total of 10 hours of class + lab where I…

My first real program was a UVEPROM copier. It was written in MC6800 Machine Code, and we had 256 bytes (not kilobytes) of RAM for everything; including the code. That was in 1983.

I am currently working in Swift, with an LLM, on a fairly good-sized app, in Xcode, for a device that probably has a minimum of 64 GB of storage, and 8 GB of RAM.

I don’t really miss the good ol’ days, to be honest. I’m having a blast.

Re: I’m spending months coding the old way

#157

Post by mswphd »

fouronnes3 wrote: post #149

I wonder if we could design a programming language specifically for teaching CS, and have a way to hard-exclude it from all LLM output. Kinda like anti virus software has special strings that are not viruses but trigger detections for testing. This would probably require cooperation during model training, but now that I think of it, is there adversarial research on LLM? Can you design text data specifically to mess w…

LLM's train in some standardized ways to emit things like tool calls, right? if you make those tokens a fundamental part of your programming language, it's possible you'd be able to run into tokenizer bugs that make LLMs much more annoying to use. Pure conjecture though.

Re: I’m spending months coding the old way

#158

Post by brianush1 »

phaser wrote: post #36

Here’s how i do it: I create a lot of stuff using AI to the max, but I also spend the necessary of time on reviewing that the AI is producing code that passes my cognitive load standards. this involves some tokens spent on grooming code and documenting well. Most of this is effortless thanks to an AGENTS.md based on this: https://github.com/zakirullin/cognitive-load/blob/main/READM... but i have a good sense of catch…

> Don’t abuse DRY, a little duplication is better than unnecessary dependencies.

That's an interesting thing to include. I agree with this point in principle, but I've found that Claude, at least, duplicates logic FAR too often and needs nudging in the other direction.

Re: I’m spending months coding the old way

#160

Post by cobbzilla »

apricot wrote: post #86

I am this very term teaching 18-year-old students 6502 assembly programming using an emulated Apple II Plus. They've had intro to Python, data structures, and OO programming courses using a modern programming environment. Now, they are programming a chip from the seventies using an editor/assembler that was written in 1983 and has a line editor, not a full-screen one. We had a total of 10 hours of class + lab where I…

I love all of this, but at least let your students use vi, it was around back then (or close). plus they don’t have to give it up when they go back in the real world, it’s an evergreen skill!
Post reply on HN