sendfoods

That makes things clearer, thanks!

Please excuse my naive question - isn't Gaussian Splatting usually used to create 3D imagery from 2D? How does providing 3D input data make sense in this context?

Which blog/post/book was this? Thanks

This looks great, thanks for sharing. I'll give it a spin on my RPI.

What a tongue-twister for English speakers, though :)

Water kefir is fun during summertime to have something a bit more healthy than soda. (Until a bottle flavored with raspberry explodes and turns your kitchen walls pink....)

Ah, I see!

2.5% of veggie weight? Or water weight?

I always use 3% of water weight, which in more consistent than weighing the veggies, because the amount of water can vary.

Very unusual to see food-related content on HN. I love it!

That's 6.5% salinity. I am wondering how you settled with that number, I have always stuck to 3% and have never seen recommendations higher than 4%! Just curious

Thank you!

May I ask what kind of blogging engine/site generator you used for the docs?

One question - how are you able to generate links to the commits, by hash?

Could you go into some details here?

- sensors used

- postprocessing

- registration algorithm(s)

Are all things that would interest me greatly :)

Great to see continued development of rerun!

I'm a SLAM engineer and work with C++ and ROS1 mainly, so I am exited to see the C++ SDK about to be released. The ROS visualization tools can be a bit crusty, even in ROS2.

That makes me unreasonably happy to hear! I really miss good mexican and latin food in mainland Europe.

Of all places I could think of spreading the word HN was definitely the last place :D

It's honestly a shame how hard it is to find true latin american foods and restaurants in parts of europe. I stuff my face with anything mexican/latin american every time I visit the States...Already miss it.

Masa Harina is becoming pretty easy to find though. Most asian (for some reason) grocers carry it.

To my knowledge, zwieback is made from wheat, not corn. /s

Not sure where you are based, but recently found these german farmers [0] that use local corn with imported machinery for their products. They also sell fresh masa (which, I guess, you are referring to?) as opposed to already pressed tortillas.

[0] https://www.tlaxcalli.de/

BBC did a pretty good video [0] about it. Haven't really heard much about it since. Must be extremely exciting for scientists, as (to my understanding) this kind of self-fertilization is completely unheard of!

[0] https://youtube.com/watch?v=CFyd-kC6IUw

I was impressed as well!

I tend to somewhat agree. This [0] repository was posted here last week and really made me think about much of what you are discussing here.

[0] https://github.com/s0md3v/roop

Thanks for sharing. How does it compare to copier [0]?

[0] https://github.com/copier-org/copier

This looks amazing, thank you for sharing.

What was the timeline, if you don't mind me asking? When and why was your wife skeptical of social media, and when did you truly start understanding her concerns?

The saga started way back with the Copperhead OS situation [1]. I believe Daniel has never been able to work through that breach of trust, and that has bled out into his behavior online and the real world.

Note it's very hard to find mainstream sources on this.

[1] https://piunikaweb.com/2019/02/05/the-demise-of-copperheados...

The whole saga is tragic. There are so many details that I don't know about, which probably makes me unable to judge the whole situation fairly. I am a very happy Graphene user, for what it's worth.

Thank you for your contribution. All the best, Daniel.

Do you know how this compares to [PrivateGPT](https://github.com/imartinez/privateGPT). I am honestly at the point of choice paralysis with all these new tools

what are you talking about exactly?

Please excuse the unrelated question...I have seen this theme multiple times in the last couple of months, and believe it is open source. Does anyone if this is the case, and where to find it? Couldn't find a mention on the site itself

I would imagine SAM being a very interesting technology for robots like this. SAM 2D segmentation to 3D projection, clustering/object segmentation based on this initial guess, then a grasping neural net.

How realistic is CPU-only inference in the near future?