shrike's recent activity
-
Comment on Ghostty 1.3.0 has been released in ~comp
-
Comment on Ghostty 1.3.0 has been released in ~comp
shrike Link ParentTech wise kitty is fucking amazing, features that were first in kitty have been ported to others, like the ability to show images in terminals, some keybinding tricks etc. But still not supporting...Tech wise kitty is fucking amazing, features that were first in kitty have been ported to others, like the ability to show images in terminals, some keybinding tricks etc.
But still not supporting the dev :)
-
Comment on Ghostty 1.3.0 has been released in ~comp
shrike Link ParentKitty is done by the same dude as Calibre and he has a "my way or the highway" attitude to programming. He does technically cool stuff, but I don't use his apps unless I have to (Calibre) because...Kitty is done by the same dude as Calibre and he has a "my way or the highway" attitude to programming. He does technically cool stuff, but I don't use his apps unless I have to (Calibre) because of his general attitude towards good programming practices.
I used to use Wezterm, because I do like applications that use Lua for configuration - but in the end it turned out to be a bit wonky. Ghostty with a few config tweaks is good for me.
-
Comment on Dox with Grok in ~tech
shrike LinkI've been pretty good at not linking from one social media site to my profile on another one and I specifically don't tell people where I live (exactly) nor the composition of my family etc.In summary, the identity remains anonymous and untraceable from public sources—it's a generic handle without unique breadcrumbs. If you have additional details from the Reddit activity (e.g., specific posts or topics), that could allow for narrower analysis.
I've been pretty good at not linking from one social media site to my profile on another one and I specifically don't tell people where I live (exactly) nor the composition of my family etc.
-
Comment on What is your top, unknown, non fiction recommendation ? in ~books
shrike Link ParentI’ve heard amazing things about this book (and the audiobook), Rory’s talks on YouTube are all fascinating dives into the human mind useful for everyone, not just marketing peoplealchemy by Rory Sunderland. About people’s expectations, perceived value and advertisement
I’ve heard amazing things about this book (and the audiobook), Rory’s talks on YouTube are all fascinating dives into the human mind useful for everyone, not just marketing people
-
Comment on The future of AI in ~tech
shrike Link ParentCrazy people are more likely to vote Rebpulican? =)Crazy people are more likely to vote Rebpulican? =)
-
Comment on What is your top, unknown, non fiction recommendation ? in ~books
shrike LinkPeopleware should be mandatory reading to anyone working in, with, or leading teams in the software industry. It's from 1987 and lays out clear step by step instructions how to have an efficient...Peopleware should be mandatory reading to anyone working in, with, or leading teams in the software industry.
It's from 1987 and lays out clear step by step instructions how to have an efficient team of programmers. It's 2026 and we're using none of them properly.
-
Comment on Is it worthwhile to run local LLMs for coding today? in ~comp
shrike Link ParentIt's been said many times, but using a frontier model for programming is like being a team lead or project manager for programmers. You tell them what you want and check the result. Either just...But yeah, agreed that my expectations were blown out of the water while working with some frontier models. Even if they stay just as they are now, this will be massively disruptive to nearly all work done in front of a computer.
It's been said many times, but using a frontier model for programming is like being a team lead or project manager for programmers.
You tell them what you want and check the result. Either just use it and see it does what it was designed to do or read the generated code before accepting it. I've shipped full features and bug fixes at $dayjob without touching a single line of code myself.
I've read a shit-ton of generated code though, my job is to deliver code I have proven to work after all and I need to be able to explain the code I'm delivering. The biggest outward change in my PRs is that there are a fuck ton of tests for each case.
When I had to type every test by hand, I'd maybe add one test that checks the specific case. Claude on the other hand will write 10-15 tests from slightly different angles including integration tests to see that the bug can't appear at all, and it won't get bored or complain.
Same goes for my personal projects, very very few had even a single test. But currently every project that moves from "hmm, would cool to do this" to me actually using them has pretty robust tests.
-
Comment on Is it worthwhile to run local LLMs for coding today? in ~comp
shrike Link ParentIt is and it isn't. The amount of money being circulated around is completely bonkers, they're just throwing billions everywhere like they can just create it out of thin air. That shit IS a bubble...It's also really got me wondering if the current AI boom really is a bubble.
It is and it isn't. The amount of money being circulated around is completely bonkers, they're just throwing billions everywhere like they can just create it out of thin air. That shit IS a bubble and will burst at some point, how epic the pop is - nobody really knows
But that just mostly means that the 20€/month plans will go away and people have to start paying market price for online AI use. Either in cash or via ads or something similar.
-
Comment on Apple announces Macbook Neo, a new budget Mac in ~tech
shrike Link ParentAnd the 500 tabs you have open are all vibrant and alive, often revisited? =)And the 500 tabs you have open are all vibrant and alive, often revisited? =)
-
Comment on Apple announces Macbook Neo, a new budget Mac in ~tech
shrike Link ParentYea, they're definitely good, not just 100% magic like the newer ones :D Took me a few years to figure out it wasn't clicking for real anymore, but rather it was "fake". The pad itself didn't...Yea, they're definitely good, not just 100% magic like the newer ones :D
Took me a few years to figure out it wasn't clicking for real anymore, but rather it was "fake". The pad itself didn't move, it just felt like it. It's even more pronounced with the external trackpad.
-
Comment on Is it worthwhile to run local LLMs for coding today? in ~comp
shrike (edited )LinkIf you want to learn how to use AI Agents, spend the 20€ or something for Claude Pro for a month and start using it. Both the app and Claude Code on the CLI. Try creating a tool with it that fixes...If you want to learn how to use AI Agents, spend the 20€ or something for Claude Pro for a month and start using it. Both the app and Claude Code on the CLI. Try creating a tool with it that fixes an issue in your day to day life or work, see how it goes. Automate something you need to do manually 50 times a week. Or make a silly game.
Now you have a baseline of what the state of the art can do
Then you can start experimenting with local models. Grab LM Studio, Ollama and ComfyUI and see what kind of free/open models there are. Some are good for coding, others can describe and even generate images.
Find the limits of the mainstream models, what will they do and what they won't. Try writing an AI assisted short story about a murderer and see how the model starts moralising on the character's actions or refuses to write about some things. Then grab some uncensored ones and get REALLY FUCKING WORRIED because they will generate detailed stories of the most heinous shit with no limits at all.
Try to recreate the same app you did with Claude with local models, it's slower of course, but how is the quality compared to it? Good enough? Try using a local model in an IDE for autocomplete or agentic workflows, how does it feel?
ComfyUI is fun too, you can easily create "pipelines" for image generation, fully locally. Also see the bits about limited and unlimited models from the paragraph about text-only models. Oof.
-
Comment on Is it worthwhile to run local LLMs for coding today? in ~comp
shrike Link ParentNo 1 depends a lot when you end the generation loop. If you don't give the AI Agent (you are using an agent and not copypasting from ChatGPT web, right?) tools to validate its work by building,...No 1 depends a lot when you end the generation loop. If you don't give the AI Agent (you are using an agent and not copypasting from ChatGPT web, right?) tools to validate its work by building, testing and linting the code - of course it's going to be shit.
I've personally done full-ass PRs and bug fixes with nothing but prompts to Claude + Opus 4.6 and the code quality is on par or a bit over what I'd write. Mostly because the model knows a few language/library tricks better than I do. Zero errors, zero hallucinations. And me writing it doesn't matter, I READ it before I submit to any human eyes - as you should. My job is to deliver code I have proven to work
AI isn't going anywhere, someone opened the Pandora's box or large language models and assossiated tech. The only thing that varies is that whether we get local models that are good enough for daily use or do we need to rely on online models.
-
Comment on Apple announces Macbook Neo, a new budget Mac in ~tech
shrike Link ParentI still don't get the idea of a touchscreen laptop. I try and try to imagine what would be a situation where I'd prefer to slather my grubby hands on my laptop display but fall short every time....I still don't get the idea of a touchscreen laptop. I try and try to imagine what would be a situation where I'd prefer to slather my grubby hands on my laptop display but fall short every time.
Those fancy foldables that can fold into a tablet, mayyybe. But a normal laptop with a hinge that doesn't go past 180 degrees. I don't get it.
-
Comment on Apple announces Macbook Neo, a new budget Mac in ~tech
shrike Link ParentThey did downgrade the trackpad from the fancy force feedback thingie to a "normal" clicky one they had in the Intel generation.They did downgrade the trackpad from the fancy force feedback thingie to a "normal" clicky one they had in the Intel generation.
-
Comment on Apple announces Macbook Neo, a new budget Mac in ~tech
shrike Link ParentThe A-series CPU in the Neo is faster in single core benchmarks than the M1. Which is kinda bonkers.The A-series CPU in the Neo is faster in single core benchmarks than the M1. Which is kinda bonkers.
-
Comment on Apple announces Macbook Neo, a new budget Mac in ~tech
shrike Link ParentMight I direct you to a new discovery in the world of Web Browsers, called BOOKMARKS :DMight I direct you to a new discovery in the world of Web Browsers, called BOOKMARKS :D
-
Comment on What are you no longer a fan of? in ~talk
shrike Link ParentWe all thought he'd be Tony Stark. Instead he is Justin Hammer.We all thought he'd be Tony Stark.
Instead he is Justin Hammer.
-
Comment on Looking for vibe-coding guides (best practices, etc.) in ~tech
shrike LinkEven if you go full Vibe, you can still ask the LLM to explain what it did in detail: https://simonwillison.net/guides/agentic-engineering-patterns/linear-walkthroughs/Even if you go full Vibe, you can still ask the LLM to explain what it did in detail: https://simonwillison.net/guides/agentic-engineering-patterns/linear-walkthroughs/
-
Comment on How do you remember? in ~tech
shrike LinkMy mind works like a vector database, I remember connections. Not things so much. I have all of my bookmarks (the stuff people keep in decade old tabs) in Raindrop This is for those things where...My mind works like a vector database, I remember connections. Not things so much.
I have all of my bookmarks (the stuff people keep in decade old tabs) in Raindrop
- This is for those things where you remember you saw that page once but can't either find it with a web search or can't remember any keywords
- These will, at some point, go to a self-hosted solution but I don't have the brainpower to go through all 5000 options.
Daily notes and stuff I need to remember goes in Obsidian
- Daily notes are for remembering "what the fuck did I do on Thursday two weeks ago"
- The other stuff is just "how do I set up a new Linux VM from scratch" or "What are the shortcuts for my Keychron keyboard's RGB lights"
- I prefer Obsidian because they now have a CLI tool AND I can use any LLM to wade through it with vague questions like "what did I say about my PT appointments last year?" or "when did I start watching Homeland?"
If something needs attention it's either in my phone calendar or Things3
- These are the stuff I need something to actively remind me about, appointments, things I need to remember to do daily (medications, workouts etc)
He's perfectly within his rights to do things like he wants to. IIRC he even wanted to rather fork Python 2 and keep a private fork than port Calibre to Python 3 :)
But I personally don't want to support that kind of behavior by using their software if I can avoid it. Calibre has no real alternatives so I just deal with the unique UX patterns and very specific ways it operates contrary to any established standards.
With terminals I have options so I choose not to use Kitty.