mark_l_watson

6 ms·

1. 61.▲
   by mark_l_watson 2mo ago

   I love the wave of new small model releases. Pleasantly surprising that an NVIDIA model runs so well on Apple Silicon using MLX! I was using nemotron-3.5-lightning:30b-mlx with OpenCode on my old (cheap) Mac this morning and no bad experien
2. 62.▲
   by mark_l_watson 2mo ago

   I used to love using Squeak but moved to Pharo to get better high resolution display support. I should try the Squeak 6.1 release.
3. 63.▲
   by mark_l_watson 2mo ago

   From the Ollama docs for this new model: Ollama's MLX engine provides state-of-the-art performance on Apple Silicon, with support for DFlash and image input: ollama run muse-glimmer:30b-mlx
4. 64.▲
   by mark_l_watson 2mo ago

   To be honest, I never give benchmarks a look. I just use the models for whatever I need to work on, so I can't really make comparisons that are useful for other people.
5. 65.▲
   by mark_l_watson 2mo ago

   Meta is rocking AI. As of last week I have been using their excellent muse coding harness with their model Muse Spark 1.2. Starting this morning I am running their new local 30B model muse-glimmer on my old MacMini 32G using Ollama (remembe
6. 66.▲
   by mark_l_watson 2mo ago

   He did the right thing. The very idea of Claude copying an existing project to answer one of my prompts or a whole project in Claude Code is a nightmare scenario. I have all but given up using the multiple trillion weight models in favor of
7. 67.▲
   by mark_l_watson 2mo ago

   I am retired, so this may not be of general interest: I am working on my fourth AI coding agent, this one written in Racket Scheme: https://github.com/mark-watson/racket-coding-agent I don't necessarily suggest th
8. 68.▲
   by mark_l_watson 2mo ago

   Are you one of the authors? Great job! It took me just a few minutes to install Jolt and build a native executable for the Chess program in my Clojure book. Jolt rocks!!!
9. 69.▲
   by mark_l_watson 2mo ago

   Clojure hosted on Chez Scheme (same as Racket). So cool. I tried building a standalone executable for the Chess game in my Clojure book using "jolt build -m chess-game.cli -o chess_jolt" and it simply worked great. I tried again w
10. 70.▲
    Jolt: Clojure compiler implemented with Chez Scheme(jolt-lang.github.io)

    213 points by mark_l_watson 2mo ago|90 comments
11. 71.▲
    by mark_l_watson 2mo ago

    I just saw this - works great to build small self contained apps. (I just tested on macOS, also supports Linux).
12. 72.▲
    by mark_l_watson 2mo ago

    I find pi-sandbox works well for me. Did you evaluate pi-sandbox? If not, give it a try.
13. 73.▲
    by mark_l_watson 2mo ago

    Same experience, sometimes using Little Coder helps make things faster because it is built on pi and strips out some functionality. Probably best to stick with pi however, adding pi-sandbox for some security. I will only run Little Coder in
14. 74.▲
    by mark_l_watson 2mo ago

    I am starting to share the opinion in the article. I used to use Claude Code with strong open models, and sometimes pay for Opus. Now one of my main drivers is out of the box pi’s 4 built in tools, and I add just two extra tools: pi-sandbox
15. 75.▲
    by mark_l_watson 2mo ago

    No, but a good idea. I want something simple, practical, understandable. I think I will write a short book on this general topic.
16. 76.▲
    by mark_l_watson 2mo ago

    Great article! I am currently writing my second harness (first was in Emacs Lisp, using Emacs as UI; second is a command line coding agent written in Common Lisp). If anyone wants to argue that it is inefficient writing your own AI coding h
17. 77.▲
    by mark_l_watson 2mo ago

    While I agree with the article, I would carry the idea further: the huge value in AI is improving ourselves. By all means use AI to help design, write tests, and implement code and while I understand this might be the big value add for corp
18. 78.▲
    by mark_l_watson 2mo ago

    Thanks, you make me feel better for spending the (fun!) time writing my first harness in Emacs Lisp with a ‘Emacs UI’ and later writing a command line harness in Common Lisp. Am I more productive with my own harnesses? Not yet, but the seco
19. 79.▲
    by mark_l_watson 2mo ago

    Odd results for me. Last month I tested five AI chat sites, with web search turned off, and four of them had a shadow of information about me as a person (what kind of books I write, what tech I use, and a random bit of other information).
20. 80.▲
    by mark_l_watson 2mo ago

    Such great advice. I have been meditating for over 50 years, and while this meditation practice is key to my being happy and content in my personal life, hobbies, and work the techniques in this excellent article are something I have only t
21. 81.▲
    by mark_l_watson 2mo ago

    You can buy really good Ramen ingredients online and make your own. If you are not a vegetarian and don’t mind eating pork, you can buy sealed packets containing about 2 or 3 fluid ounces of thick pork bone broth, reduced to a thick liquid.
22. 82.▲
    by mark_l_watson 2mo ago

    Are you using the $20/month subscription? I don't use it much, perhaps two 20-minute sessions a day so I don't get close to using my quota. When it works, it is a great system, but now three times in a week it wasted some of
23. 83.▲
    Ask HN: Has Claude Code been (mostly) usable for you?

    3 points by mark_l_watson 2mo ago|6 comments
24. 84.▲
    by mark_l_watson 2mo ago

    This is my experience also: DeepSeek v4 flash is good enough for most of my work and I like the fast response times. I buy tokens mostly from FireWorks.ai in the US, but I also prepaid for a large chunk of tokens directoy with DeepSeek. I u
25. 85.▲
    by mark_l_watson 2mo ago

    I toggle back and forth between deepseek v 4 flash/pro on FireWorks.ai using OpenCode. Easy to toggle, I default to flash.
26. 86.▲
    by mark_l_watson 2mo ago

    I prepaid for a ton of DeepSeek v4 flash tokens. When I use them I would like to try purchasing Luna tokens for a while. I like cheap and fast models and since I am retired, if I waste time having to sometimes manually switch to a stronger
27. 87.▲
    by mark_l_watson 2mo ago

    > I've found that I get more done the less time I'm at the computer. Right on! For 15 years I worked for two companies that had their offices on the cliffs overlooking the ocean in La Jolla CA. I would spend a lot of time walki
28. 88.▲
    by mark_l_watson 2mo ago

    Very good, pairs well with the linked article. I am an old man (mid-70s) and retired now except for personal research and writing, but fifty years in the business taught me to simply pay attention to what people actually get done.
29. 89.▲
    by mark_l_watson 2mo ago

    This looks really good, but for me the timing of seeing agent-manager is bad: I read about the similar Herdr on HN several days ago and I have completely changed my dev setup for the better. I have always been a traditional simple command l
30. 90.▲
    by mark_l_watson 2mo ago

    +1 upvote I don’t agree with you but your comment is very well thought out and doesn’t deserve downvotes.