ashish01

Since it was making rounds yesterday here is what muse generated for the 3d rubik's cube prompt - https://ashish01.github.io/rubik-muse/

Browser not fully supported

This editor relies on WebGPU for rendering, which currently works best in Chrome or other Chromium-based browsers. You may experience issues in your current browser.

(reminds of works in IE6)

That is really beautiful literate program. Seeing it after a long time. Here is a opus generate version of this code - https://ashish01.github.io/microgpt.html

Still working on PocketWise (https://pocketwise.app), a simple double-entry accounting app. Just finished adding end-to-end encryption with a zero-trust server model. All encryption and decryption happens in the browser (using PRF), and the server only sees encrypted data.

I have an old Lenovo IdeaPad with fairly modest hardware, and I have both Fedora and Windows installed. About 90% of the time I use Fedora, and it works fine overall. The only thing that bothers me is that Firefox on Fedora feels noticeably more sluggish compared to Edge or Firefox on Windows. Maybe it’s just a perception issue, but I’d love to know what others are using as their web browser on Linux.

Simple meal tracker to give some macros but mainly give a health rating on 1-10 scale. http://meal-tracker.fly.dev

I use Litestream for near real-time backups. Does not change how SQLite is used on the server, just a replacement for .backup

I am working on PocketWise (https://pocketwise.app) a lightweight personal finance tracking app. Goal is to make double entry accounting simple and approachable for everyday use. It’s my first project of this kind, so I’d really appreciate any feedback.

https://hn-reader.pages.dev/

I also prefer BFS expansion of comments. So I implemented https://hn-reader.pages.dev/ for this. Also has dark mode to be easy on eyes.

one of the biggest waste of money on an acquisition.

I think that was when intel acquired McAfee for 8B in 2010.

I wrote one a while back https://github.com/ashish01/hn-data-dumps and it was a lot of fun. One thing which will be cool to implement is that more recent items will update more over time making any recent downloaded items more stale than older ones.

With a one-time setup, a low-fi solution is to receive an email from your bank for every transaction and extract the data into a ledger entry.

You do get $10 worth of monthly claude credits for free - https://zed.dev/blog/zed-ai-billing but for others like OpenAI or DeepSeek you need an API key.

https://github.com/egh/ledger-autosync is really useful for integrating banks OFX files into an existing ledger file. Over time, it learns common patterns, automatically assigning transactions to the correct accounts.

Is there a way to download the problems via an API, or have them available in a file like https://norvig.com/top95.txt?

It would make a great dataset for benchmarking different solver approaches.

I agree with the sentiment of the blog, but it doesn't present a very compelling argument. Just relies on authority rather than evidence or logical reasoning. For a more impactful and well-reasoned discussion on this topic checkout "Simple Made Easy" by Rich Hickey (https://www.youtube.com/watch?v=SxdOUGdseq4)

I made two using scd41, esp32 and esphome. https://shop.m5stack.com/products/co2l-unit-with-temperature...

I tried jit as well but looks like recursive functions cannot be jitted yet :(

I was curious about how slow (or fast) it is compared to cpython. On fibonacci.py rustpython about 11x slower than cpython.

    def fib(n):
        if n == 0 or n == 1:
            return 1
        return fib(n-1) + fib(n-2)

    print(fib(35))

    time python3 ~/code/fibs.py
    14930352

    ________________________________________________________
    Executed in    1.18 secs    fish           external
       usr time    1.14 secs  180.00 micros    1.14 secs
       sys time    0.01 secs  616.00 micros    0.01 secs

    time ./target/release/rustpython ~/code/fibs.py
    14930352

    ________________________________________________________
    Executed in   13.44 secs    fish           external
       usr time   13.32 secs  175.00 micros   13.32 secs
       sys time    0.02 secs  776.00 micros    0.02 secs

Let Me GPT That For You! -> https://gprivate.com/65ubk

Any good examples or notebooks using RL to solve typical optimization problems?

PuLP is a great way to get started, and I used it as my first attempt for many problems. However, I have since switched to using Google OR tools (https://developers.google.com/optimization/pack/bin_packing) or Python's Z3 interface, depending on the problem I'm working on.

One of the core problems in semantic search is doing efficient approximate nearest neighbor search. In this post the author uses KD trees. There are other alternatives for efficient ANN like scann from Google and faiss from Facebook.

Any books / papers you would recommend to cover basics of search engine implementation? Practicalities like elias fano encoding and WAND searching.

I wrote a simple hn data downloaded based on this api. Last I remember I was able to download the entire corpus by letting it run overnight.

https://github.com/ashish01/hn-data-dumps

This brings back so many memories. I was in a same situation (reading a book) and needed on digital offline dictionary. I couldn't find anything for my phone back then. It was a Sony Ericson phone with 64MB memory and J2ME as development environment.

So I wrote a offline dictionary with a friend of mine based on wordnet and J2ME polish. It was my first real world project - http://cornucopia.sourceforge.net/

A chat server

If you are trying out parsing in python, I would recommend giving [Lark](https://github.com/lark-parser/lark) a try. I have used it for smallish projects are its really easy to use especially with the Earley parser.

I tried a few QnA style queries and most (all?) of them got their answers from Wolfram Alpha.

why is sky blue

Top result: https://www.wolframalpha.com/input/?i=why%20is%20sky%20blue

what is the capital of punjab

Top result: https://www.wolframalpha.com/input/?i=what%20is%20the%20capi...

what is the population of seattla (with typo)

Top result: https://www.wolframalpha.com/input/?i=what%20is%20the%20popu...

For one keyword query I tried "rust lang" the results felt very repetitive.

Just wondering if its a layer on top of Wolfram Alpha or is it doing QnA inference on its own.