Really, truly blown away. I'm sure there are plenty of edge cases to correct for but I haven't gotten so excited by a demo in a long time. I've obviously signed up to learn more.
Really, truly blown away. I'm sure there are plenty of edge cases to correct for but I haven't gotten so excited by a demo in a long time. I've obviously signed up to learn more.
First, a clarification: Since people have brought up the topic of "hard to believe one guy handled all aspects", I want to state that while this is my concept and my project (and I've devoted nearly every waking hour since 2020 to it), I'm no jack of all trades.
In the last two years, for example, I've gotten occasional but much-needed help (and knowledge) from two freelancing developers for some aspects that I don't have expertise with, including Linux intricacies, DMA-based firmware programming, UI design details, and some other subtleties. A more experienced PCB designer (colleague from past) has also helped me, particularly with the recent 6-layer PCB layouts. Plus, an assistant in the past has aided with the soldering of some tightly packed boards.
Last but not least, my SO has contributed graphics + Adobe Premiere effort for my videos; her equally important contribution was being a frequent listener, and sometimes a much-needed boundary, to my evolving thoughts through this project's journey.
As the project evolves to its next phase, I do hope to get more people into primary development of this modular ecosystem, in the form of both team members and eventually custom Block designers from the community, once I organize and release the necessary files + documentation for everyone to work with.
I'm perhaps late to the party, but I'll try to answer as many questions as I can on this thread now, including hopefully the ones posted several hours ago!
How are you planning to commercialise this?
Awesome and Inspiring work!
Contact in the HN profile.
The algorithm that chooses the most likely/useful application for infinite possible combinations of modules would be an impressive feat for a single person.
The multidisciplinary coding required to get all of these pieces to work together in a hot-swap way would be an impressive feat for a single person.
The design and implementation of a single physical connection interface that can adapt or carry all these different protocols (USB, HDMI, etc) would be an impressive feat for a single person.
Not to mention the PCB design, 3D enclosure design, machine learning proficiency, aesthetic product design chops, and on and on.
All together, this is unbelievable. This is 0.01% level stuff. Mozart, Musk, Melville. Somewhere in that neighborhood.
However, the most complex and costly part is the PCB and circuit design. The PCBs used in the blocks are absolutely awesome.
Thanks to 3D printers that are cheap, the casing and other plastic materials are easy to make.
All that being said, this has very low chance of becoming a real world product. Real world is messy, dirty, wet and an absolutely shitty place for snap on electronics.
What would work, is better connector technology. It is obvious that even with all this simplification, this will still be a hobbyist product, rather than a serious mass market product.
> Merveilles seems aligned with the ideals of Solarpunk while internally expecting the world of Cyberpunk, it is neither a utopian or dystopian vision, but a way of straddling both contingencies.
Two things stick out for critique to me -
> Merveilles seems aligned with the ideals of Solarpunk while internally expecting the world of Cyberpunk, it is neither a utopian or dystopian vision, but a way of straddling both contingencies.
I get this, and in a way, I think it’s how I’m operating already, but man, it’s an art movement - don’t give the dystopia space in the room, it’s already got plenty everywhere else.
> The Merveilles visual aesthetic restricts color palettes to black and white, vector or pixel art, with at most a single accent color (usually a sea-foam aqua). Industrial design is minimalist, geometric black-forged metals, natural wood.
My visual aesthetic these days is “all of the above.” For the love of god, colors exist - trillions of them! Take two! Hell, take three or four! They’re cheap! And shapes - my god, man, the shapes you can make! You ever see the temple carvings in Nepal? So many shapes! Take a walk through a forest, and just look at all the shapes! Look at trees, man - the opposite of simple!
(this was intended as a gentle ribbing, not actual criticism. I found the contrast between exclusion and inclusion interesting :)
https://www.fsf.org/news/artist-collective-hundred-rabbits-t...
So likely they won't have nearly the cycle life of USB-C but do they really need to?
Most user-facing connectors have a metal shield around them connected to frame ground. The idea is that any ESD shock goes safely through this first instead of to a sensitive data line.
Connection-sequencing ensures that the ground and power is connected before data lines are. If you look inside a cable-side USB connector, you'll see that the inner two wires (data) are recessed a little so that the power connectors go first. A device that is connected without power can (through its ESD protection circuit) attempt to draw power from data lines... this can cause damage because most data lines can't supply the current to power the entire device.
Every USB device connected to your PC right now probably has a little ESD protection IC in it. Usually sitting right next to the USB connector (as close as possible).
And since the snap-in boards are all rigid you don't really need to worry about the possibility that contact will be made in like.. a rolling way. Either the pins are on the board or they're not.
But I'm not an EE, just a hobbyist, so maybe there's something I've missed there and what I'm thinking is way more complicated than I think.
Google tried this in 2015. It was called Project Ara.[1] A similar click-together system, with the same problems. Plus the problem that there's no good reason to make a phone that way. Google acquired that from Phonebloks and killed it.[2]
It's not a bad idea if you solve the mechanical problems and have the components produced in sufficient quantity to be very cheap. A little bigger, a lot more rugged, with a good wire management solution, and you'd have something useful.
[1] https://www.cnet.com/tech/mobile/project-ara-everything-we-k...
It is briefly explained at 4:04 that the initial 4 mins it's running on a 3300 battery, and then he swaps it to USB charging.
Having the home automation communications be a message queue instead of a general-purpose communication protocol is so much better.
As solo founder this is really motivational - I hope my upcoming project garners even half the interest once launched! So many hats.