KronisLV

50 ms·

871.
by KronisLV 1y ago
> For example you know context and a poor explanation of what you are after. Googling will take you nowhere, LLMs will give you the right answer 95% of the time. This works nicely when the LLM has a large knowledgebase to draw upon (form
872.
by KronisLV 1y ago
Hey, thanks for working on the software, you made something really cool!
873.
by KronisLV 1y ago
> yes, it's a cartoon character; get over it I think it's cute. It wouldn't be a problem if it wasn't cute. It's inoffensive and doesn't get in the way of the software serving its function, the only critique
874.
by KronisLV 1y ago
Typically you want your audit/log data to be immutable and kept in an append only data store. Whether that's a typical relational DB or something more specialized (like a log shipping solution) that's up to you, but usually i
875.
by KronisLV 1y ago
> And the problems are significant: > High vendor lock-in – When GitHub wrote Git LFS, the other large file systems—Git Fat, Git Annex, and Git Media—were agnostic about the server-side. But GitHub locked users to their proprietary se
876.
by KronisLV 1y ago
> Schema design should be flexible, because once you have thousands or millions of records, it can be an enormous pain to change the schema. However, if you make it too flexible (e.g. by sticking everything in a “value” JSON column, or u
877.
by KronisLV 1y ago
It's interesting that mod_md is so unknown: https://httpd.apache.org/docs/2.4/mod/mod_md.html But also hey, now we have built-in ACME support in all the mainstream web servers: Nginx, Caddy and Apache2!
878.
by KronisLV 1y ago
> And one of the best parts is, I've never had a Debian system break without it being my fault in some way. https://blog.kronis.dev/blog/debian-updates-are-broken https://blog.kronis.dev/blog&#x
879.
by KronisLV 1y ago
I did 20 questions. In 75% of the answers I picked GPT-5, that's a pretty strong result, at least when it comes to subjective preferences!
880.
by KronisLV 1y ago
> Are there any games that teach automation like Factorio but that don't have that depressive dystopian magnasanti feel? And as the opposite question: are there games that give more of that feeling? I want to feel like I'm pl
881.
by KronisLV 1y ago
> Chat control is bad, but it was voted out multiple times already, and is fundamentally incompatible with several member states' constitutions (based on a comment, didn't read more into it, but I believe Romania and Germany, a
882.
by KronisLV 1y ago
My current GPU is an Intel Arc B580 and before then it was the A580. They aren't perfect but now that they're coming down to MSRP, they're pretty good budget friendly options. I would have also unironically gotten a Core Ultr
883.
by KronisLV 1y ago
How hard would it be for the model to be like: "Okay, there's a bar chart in the picture, the left bar is 350 px in size and the right bar is 120 px. Meanwhile, the labeled values are X and Y, which doesn't seem to match thos
884.
by KronisLV 1y ago
For comparison, at work we got a pair of Nvidia L4 GPUs: https://www.techpowerup.com/gpu-specs/l4.c4091 That gives us a total TDP of around 150W, 48 GB of VRAM and we can run Qwen 3 Coder 30B A3B at 4bit quantization w
885.
by KronisLV 1y ago
> Meanwhile, i found that a very good workflow is asking claude code to come back with clarifying questions and then a plan, before just starting to execute. RooCode supports various modes https://docs.roocode.com/basic-u
886.
by KronisLV 1y ago
For small enough codebases, that seems like an inevitable reality, eventually. If you have nearly limitless compute to throw at an issue and a good enough model, then it should be able to create enough test cases to cover most aspects of
887.
by KronisLV 1y ago
I've used self-hosted GitLab a bunch at work, it's pretty good there still. In my opinion GitLab CI is also a solid offering, especially for the folks coming from something like Jenkins, doubly so when combined with Docker executo
888.
by KronisLV 1y ago
> Here's my reason why I think it does produce low quality code; because it does. I’d say that the original claim is so context dependent that it bothers on being outright wrong. That’s like saying that Java sucks because I’ve seen
889.
by KronisLV 1y ago
> Someone at some point had a working mental model of the code and a reputation to protect This isn’t always true in absolute terms. In many places and projects it’s about doing a good enough job to be able to ship whatever the boss said
890.
by KronisLV 1y ago
> Aider is much nicer, from the standpoint that you can add the files you need it to know about, and send it off to do its thing. As a user, I don't want to sit there specifying about 15-30 files, then realize that I've missed
891.
by KronisLV 1y ago
I feel like sooner or later the standard for these types of discussions should become: > "Here's a link to the commits in my GitHub repo, here's the exact prompts and models that were used that generated bad output. This e
892.
by KronisLV 1y ago
Linux Mint with Cinnamon (or XFCE before that) was my daily driver for a while, really nice distro that just works and gets out of your way for the most part. I do personally find the UI of LibreOffice to be good enough (you can even change
893.
by KronisLV 1y ago
> The point of the Coffee Beans Procedure is this: if you can’t answer those questions, if you don’t even find them interesting, then you should not open a coffee shop, because this is how you will spend your days as a cafe owner. If we
894.
by KronisLV 1y ago
I’d very much prefer more padding between the clickable UI elements on mobile in particular, because the zoom in -> click upvote -> zoom out, or the click downvote by accident -> try to unvote -> try to upvote again, well, it ge
895.
by KronisLV 1y ago
I mostly use RooCode nowadays, which works well enough with both Claude and Gemini and other models, even locally hosted ones. Decoupling the LLM vendor from the tools might miss out on some finer features, but also gives me a little bit mo
896.
by KronisLV 1y ago
I personally take an issue with them expecting that your usage would be more or less consistent throughout the month. Instead, I might have low usage throughout most of the month and then an 11 hour binge a few days, which in most cases wou
897.
by KronisLV 1y ago
I really like the idea of finding what dials can be turned and then doing so and seeing how much headroom you normally have and how far away from things breaking you are. Plus, those sorts of artificial limitations can be really helpful i
898.
by KronisLV 1y ago
There’s also RooCode which is pretty nice: https://marketplace.visualstudio.com/items?itemName=RooVeter... (fork of Cline, that one’s also good) Ofc some might prefer the pure CLI experience, but mentioning that because it
899.
by KronisLV 1y ago
> SPAs make sense when your users have long sessions in your app. SPAs also make sense when you want to decouple the front end from the back end, so that you have a stable interface like a RESTful API and once AngularJS gets deprecated y
900.
by KronisLV 1y ago
> Mainstream platforms accepting Bitcoin as an option is fine and great, but few if any are going to want to only accept Bitcoin. For all of the talk (hype) about how crypto has the potential to avoid the exact type of meddling and manip

More ›