Hacker Newsnew | past | comments | ask | show | jobs | submit | d3k's commentslogin

I see a dark pattern in suburban France. Every house with kids has its trampoline, and often a play-frame and a pool too. Now, kids play with those a bit from time to time, but they rarely go to a friend's house, because they have their own thing there. So if there is a park, a pool or a trampoline in a park nearby, I would rather use them. I believe it is better for social and environmental reasons.


I see the opposite pattern here (Pittsburgh, PA). The homes with trampolines that I have seen end up being a magnet for the neighborhood kids and become a social scene of sorts.

When our children were younger, every day they'd end up playing on the trampoline with some different subset of the dozen or so other neighborhood children (who didn't own a trampoline). There was tons of laughter, some tears and conflict - and overall an amazing way for these children to spend time together as they grew up.

Our kids have aged out of the trampoline now. But I was just driving through our neighborhood yesterday and I saw two other homes with trampolines, both filled with five or six kids, rolling around, playing, bouncing - exactly how the OP described and exactly what we saw in our own backyard.


It's not just France. This is the standard pattern of suburbs, even if you don't have these things at your house.


Seems nice! I have been using Handy https://handy.computer/ for a while and I am happy with it (using Parakeet), this seems quite similar in terms of features, except that is uses Apple's APIs.


Big fan of Handy as well. I often mix languages (german/english) and Apple onboard dictation is not working well in these cases. Having the ability to switch models so far has proven to be helpful. I guess Handy should implement the new native Apple APIs though.


AFAIK, only CoreML can use Apple's NPU (ANE). Pytorch, MLX and the other kids on the block use MPS (the GPU). I think the limitations you mentioned relate to that (but I might be missing something)


I totally agree with the post. The definition of a museum is "an institution dedicated to displaying or preserving culturally or scientifically significant objects", according to Wikipedia. Most of the time I do not see anything significant on these screens in museums, since equivalent content can be easily reached on any phone. Real, relevant objects are much harder to find and find a way to create interest around them. But that is exactly what makes a museum a good museum, not the screen.


Very nice! I wish they were more keen to contribute to AI/ML community an publish weights and model definition on HuggingFace. Funny enough I have just seen today a similar demo that is using a freely available VLM: https://github.com/ngxson/smolvlm-realtime-webcam


SmolVLM is from huggingface team

cool to see people doing stuff with smaller models

https://huggingface.co/blog/smolvlm

https://arxiv.org/abs/2504.05299


Nice! it reminds me of a very similar project a friend of mine did like 10 years ago https://github.com/simias/gb-rs


Brainchip | Software Engineer | Full-time | On-site (hybrid) | Toulouse, France

Brainchip builds edge AI on-chip processing and learning IPs. We are building the next generation of our edge AI hardware and software products.

We are hiring for these roles in our Toulouse based office:

* C++ software engineer with 2+ years of experience, possibly with embedded systems. Full description: https://www.linkedin.com/jobs/view/3653985060/

* Python software engineer with experience with TensorFlow/Keras, Pytorch, ONNX. Full description: https://www.linkedin.com/jobs/view/3653973926/


It seems nice. When programming envs tend too complicated though, I tend to switch to docker: a given image will handle a given version and it's mostly hassle free. That probably would not make sense on macOS or Windows, if you want to run native versions.


Same. Getting a python environment up and running can be a multi-day endeavor if you're unlucky. Just using docker compose on the existing Dockerfile for the project and mounting the local filesystem, means everything just works. PyCharm now works nicely with a remote interpreter inside docker compose.

I've even tried doing that with all the devtools we have. Like normally a developer here has to install gcloud, terraform, kubectl, apt-get lots of stuff, and then 3-4 internal tools. Then configure all of it, and of course some stuff will just not work on someone's computer. But creating a docker image including all that means that I can grab a new computer and be up and running in minutes.


Hijacking this: does anyone know how to get vscode to use a python interpreter inside a docker container for linting? I’ve resorted to using the docker container, but also having a python env so that I can tell vscode to point to the python interpreter in the env to get linting. Otherwise, dependencies come up unresolved.

I did some googling, but I’m unfortunately not experienced enough to be able to describe the problem succinctly to google.

Using pylance, btw.

TIA!


Have you tried VSCode's remote abilities? Should do exactly as you describe. https://code.visualstudio.com/docs/remote/remote-overview or https://devblogs.microsoft.com/python/remote-python-developm...


I’ve seen these and got it to work this way. I’m not too fond of working inside the container as it’s just a little more friction, but I’ve been working this way for a little bit now. I was hoping to avoid working in the container.


Try this: 1. Bring up command palette using `Ctrl+Shift+p` 2. Search and select "Python: Select Interpreter" 3. Select your Python environment.

This works in VSCode over SSH perfectly fine - the locally running VSCode is able to detect and select Python environments in the remote server.


The trouble with this is that the python environment is in the container - In step 2, id need to type path to the interpreter that it inside the container.

Although, now you’ve got me thinking what would happen if I mounted the interpreter inside the container to something local that I can point to? Might try this later, thanks for the help!


I regularly use asdf-vm to manage kubectl, python, nodejs, and a bunch of other stuff.

"Setting up a Python environment" takes about 10 minutes and it's dead nuts reliable.

It is a very effective package manager.


This. Docker not only solves the language-specific venv problem, but also encapsulates other system-level dependencies that may not be covered by stuff like pyenv or nvm. Plus with docker-compose you can trivially satisfy other, way more complicated dependencies like "fully functional PostgreSQL DB is accessible over the network", and you can deploy the entire thing to production with a few simple commands.


One thing that I realized after finally dabbling with Python after years in other languages was just how much Python needs Docker in development. There are so many details to get the right version of python working on a project because of system level assumptions that using Python without Docker is a huge pain.

Conversely, the experience of using Ruby within Docker for development is often clunky and awkward by comparison.


Installing dependencies, per the docs, before installing python works well. Installing library dependencies for packages also works well. It's not dissimilar to compiling c/c++ where you have to read the installation docs. And the docs better be updated.


I've not found this; pyenv and poetry sort it all out for me, and I could simplify and just make do with one.


This is a common thing in software too I fear. I have Firefox on macOS. Everything is configured for US English, but I live in France. When I tell Firefox to save a page (or download something), the dialog prompts me to "Enregistrer sous..." (i.e.: "Save As..."). No idea why this happens.


It should be possible to see if a given "secure" binary makes use of a given microcode instruction. If that is the case, it could be possible to use udbgrd/udbgwr to modify the microcode instruction to do something else, e.g. to store some register data in the URAM or the SRAM. By doing this you should be able to "spy" the data the "secure" binary is processing.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: