almostdigital

https://elevenlabs.io/sound-effects

With the prompt "WWII Plane Japan Kawasaki Ki-61 flying by, propeller airplane" and setting looping on and 30 sec duration manually instead of auto (the duration predictor fails pretty bad at this prompt, you need to be logged in to set duration manually) it works pretty well. No idea if it's close to that specific airplane though it sounds like a ww2 plane to me though.

Super disappointed it didn't have Chūō Line, had to go find them :)

https://www.youtube.com/watch?v=1nkEfZKMEnw

I saw some people report that their dishes are pointing in unusual directions so it would explain that, and that my service is fine since I only use the polar orbiting ones ¯\_(ツ)_/¯

https://www.starlinkmap.org

I'm posting this from my Starlink connection, been working fine here. I'm close to the arctic circle so maybe the polar orbiting satellites are not affected

I thought those pans that are marketed as "granite" with a speckled pattern where just PTFE coated aluminium, is this something different?

Just whatever you get by default with pip install numpy... Changing the benchmark to run with a 1024x1024x1024 matrix instead of a 128x128x128 does speed up numpy significantly though

    Python           119.189 GFLOPS
    Naive:             6.275 GFLOPS            0.05x faster than Python
    Vectorized:       22.259 GFLOPS            0.19x faster than Python
    Parallelized:     50.258 GFLOPS            0.42x faster than Python
    Tiled:            59.692 GFLOPS            0.50x faster than Python
    Unrolled:         62.165 GFLOPS            0.52x faster than Python
    Accumulated:     565.240 GFLOPS            4.74x faster than Python

    Build Dependencies:
      blas:
        detection method: pkgconfig
        found: true
        include directory: /opt/arm64-builds/include
        lib directory: /opt/arm64-builds/lib
        name: openblas64
        openblas configuration: USE_64BITINT=1 DYNAMIC_ARCH=1 DYNAMIC_OLDER= NO_CBLAS=
          NO_LAPACK= NO_LAPACKE= NO_AFFINITY=1 USE_OPENMP= SANDYBRIDGE MAX_THREADS=3
        pc file directory: /usr/local/lib/pkgconfig
        version: 0.3.23.dev
      lapack:
        detection method: internal
        found: true
        include directory: unknown
        lib directory: unknown
        name: dep4364960240
        openblas configuration: unknown
        pc file directory: unknown
        version: 1.26.1
    Compilers:
      c:
        commands: cc
        linker: ld64
        name: clang
        version: 14.0.0
      c++:
        commands: c++
        linker: ld64
        name: clang
        version: 14.0.0
      cython:
        commands: cython
        linker: cython
        name: cython
        version: 3.0.3
    Machine Information:
      build:
        cpu: aarch64
        endian: little
        family: aarch64
        system: darwin
      host:
        cpu: aarch64
        endian: little
        family: aarch64
        system: darwin
    Python Information:
      path: /private/var/folders/76/zy5ktkns50v6gt5g8r0sf6sc0000gn/T/cibw-run-27utctq_/cp310-macosx_arm64/build/venv/bin/python
      version: '3.10'
    SIMD Extensions:
      baseline:
      - NEON
      - NEON_FP16
      - NEON_VFPV4
      - ASIMD
      found:
      - ASIMDHP
      not found:
      - ASIMDFHM

Here's the same benchmark with np.matmul instead of native python (on M2 MBP)

    Python             4.216 GFLOPS
    Naive:             6.400 GFLOPS            1.52x faster than Python
    Vectorized:       22.232 GFLOPS            5.27x faster than Python
    Parallelized:     52.591 GFLOPS           12.47x faster than Python
    Tiled:            60.888 GFLOPS           14.44x faster than Python
    Unrolled:         62.514 GFLOPS           14.83x faster than Python
    Accumulated:     506.209 GFLOPS          120.07x faster than Python

I did just that, it's 120x faster than numpy. See my comment to OP.

Here's the same benchmark with numpy instead of native python (on M2 MBP)

    Python             4.216 GFLOPS
    Naive:             6.400 GFLOPS            1.52x faster than Python
    Vectorized:       22.232 GFLOPS            5.27x faster than Python
    Parallelized:     52.591 GFLOPS           12.47x faster than Python
    Tiled:            60.888 GFLOPS           14.44x faster than Python
    Unrolled:         62.514 GFLOPS           14.83x faster than Python
    Accumulated:     506.209 GFLOPS          120.07x faster than Python

Haha! This is great! Maybe it could be less jarring if you put a low pass filter on the angle changes

Yeah, just realized that and was about to edit my comment. Found an issue tracking the black replacement bit: https://github.com/charliermarsh/ruff/issues/1904

This is great, I'm going to switch to this from black. Being used to working in other languages I feel like I'm swimming in molasses when using Python.

It's funny that all good things for Python are not written in Python. Says a lot about the language.

I'm offended! I don't click around like some lowly peasant, I navigate by keyboard like God intended.

In theory it seems neat but in practice I'm here with 10 tabs all called "+page.svelte" or "+page.server.ts" and it completely breaks my workflow since I can't tell them apart or navigate with fuzzy name matching. How do you deal with that?

SvelteKit's routing pattern is anything but clean. It uses filesystem based routes where every file is called "+page.svelte" (or +page.server.js/ts for API only routes).

For anything but a demo app with just a couple of routes it's a headache to navigate.

Edit: Another smell of the routing system to me is that you need to resort to complexity like this [1] to have anything but top down inheritance. I love Svelte btw and have used it in many projects.

[1] https://kit.svelte.dev/docs/advanced-routing#advanced-layout...

Looking forward to trying this, VSCode is great but I really miss the performance of Sublime Text. I hope they get the plugin system right, killer feature would be if it could load VSCode plugins (incredibly hard to pull off, yes)

Code from the paper:

https://github.com/trailofbits/polytracker https://github.com/trailofbits/polyfile

Automatic dark/light mode tracking sunset/sunrise is the best of both worlds IMHO.

This standard isn't part of the metric system though

Camera phones killed it but generative AI models will revive it.

Who cares if someone takes a photo of you on the beach when they can just as well ask the AI to see you nude in any situation they can dream of... That, or the burka will become very popular

Love this blog, I get the "Dat Parallax" achievement every time I visit :D

Another interesting idea: Drop asteroid pieces from space into the ocean to increase biomass growth

https://www.youtube.com/watch?v=2zQb_OitsaY

Link to non-paywalled paper: https://www.biorxiv.org/content/10.1101/2020.12.03.410399v4

It's interesting that Lex is so polarizing, I personally find him a great host.

Him almost never doing research is objectively false though. He often reads books/articles/tweets written by the guests and asks about them for example.

I can't recommend docker swarm enough, it does 95% of what you need in k8s without requiring someone full-time just to manage it.

Thank you!

Technical writeup here: https://lexman.rocks/about

I wish davinci-002 wasn't so damn expensive to use :(

Haha love it!

The SVG posters are cool, if you're looking for some feedback try to find some way to reduce the lag on the site when previewing them. In Safari on a M1 the site slows to a crawl once the poster SVG is loaded, each click takes several seconds to register. Maybe you could do a low res server side rendered version? That could you let users preview their repos without them being able to just right-click the SVG as well