LiamMayfair

Writing tests is the one thing I wouldn't get an LLM to write for me right now. Let me give you an example. Yesterday I came across some new unit tests someone's agentic AI had written recently. The tests were rewriting the code they were meant to be testing in the test itself, then asserting against that. I'll say that again: rather than calling out to some function or method belonging to the class/module under test, the tests were rewriting the implementation of said function inside the test. Not even a junior developer would write that nonsensical shit.

The code those unit tests were meant to be testing was LLM written too, and it was fine!

So right now, getting an LLM to write some implementation code can be ok. But for the love of god, don't let them anywhere near your tests (unless it's just to squirt out some dumb boilerplate helper functions and mocks). LLMs are very shit at thinking up good test cases right now. And even if they come up with good scenarios, they may pull these stunts on you like they did to me. Not worth the hassle.

If kids have learned to run their own Minecraft private servers, hosting a VPN should be child's play... Pun maybe intended.

I tried GPT-5 to write some code the other day and was quite unimpressed with how lazy it is. For every single thing, it needed nudging. I'm going back to Sonnet and Gemini. And even so, you're right. As it stands, LLMs are useful at refactoring and writing boilerplate and repetitive code, which does save time. But they're definitely shit at actually solving non-trivial problems in code and designing and planning implementation at a high level.

They're basically a better IntelliSense and automated refactoring tool, but I wouldn't trust them with proper software engineering tasks. All this vibe coding and especially agentic development bullshit people (mainly uneducated users and the AI vendors themselves) are shilling these days, I'm going nowhere near around.

I work in a professional software development team in a business that is pushing the AI coding stuff really hard. So many of my coworkers use agentic development tools routinely now to do most (if not all) of their work for them. And guess what, every other PR that goes in, random features that had been built and working are removed entirely, so then we have to do extra work to literally build things again that had been ripped out by one of these AI agents. smh

In certain translations of the anime, they're referred to as cousins. Quite the cousins they are

This hits the nail right on the head. The point of cloud services is to take away all the overheads of building and delivering software solutions that have nothing to do with the actual business problem I'm trying to solve.

If I want to get a new product to market, I want to spend most of my time making my core product better, more marketable, more efficient. I don't want to divert time and resources to just keep the lights on, like having to hire a whole bunch of people whose only jobs is to provision and manage servers and IT infrastructure (or nurse a Kubernetes cluster for that matter). Managing Kubernetes or physical tin servers is not what my business is about. All this tech infrastructure is a means to an end, not the end itself.

That's why cloud services is such a cost efficient proposition for 98% businesses. Hell, if I could run everything using a serverless model (not always possible or cost effective) I'd do it gladly.

I was using the Signal "notes to self" too when taking notes during talks and conferences. Taking quick pictures of the slides in context was also a key thing for me. Exporting these unstructured notes into a useful notes archive is a pain as you say, especially if there is media too.

I caught myself doing this so often that I ended up building myself an app for this specific workflow. It's rather simple, just an MVP if you will, but it works well for me. Taking notes works exactly like Signal's "note to self" but it has some QoL stuff on top of that like separate notebooks and exporting notes and pictures to a single PDF archive. I can then import the PDF archive into Notion, which is my main notes repository. Notion can now parse PDF files and import them as regular Notion pages, which closes the loop for me rather nicely. YMMV ofc

I haven't published it to any app stores yet (might do in the future) but the source code is available here if you're technically savvy and happy to build and install it yourself.

I didn't know this. That's so wholesome and adorable!

electrons be vibin

QQ what do you guys think would be the last straw for her? I'd love to know all the deets

kthxbye

snake_case_ftw

Yeah, but there are degrees of vulnerability. Otherwise, things like password strength or MFA wouldn't matter.

If all your passwords are fully random, then that's one less weakness that can be exploited. People can't make educated guesses about your passwords just from analysing your social media profiles and history, e.g. if you post a lot about Star Wars, it's more likely your passwords could contain a Star Wars reference.

That system is vulnerable to social engineering attacks. If hackers found out all their favourite things that lead to the core part of the password, guessing the prefix wouldn't be that hard. Also, what would your friend do if one of these passwords got compromised and had to change it? Would he just add a 1 to the site-specific part of the password?