I'm surprised they didn't consider hosting their own vector tile server, instead of painstakingly rasterizing their entire dataset and optimizing it over and over to fit into the Mapbox data caps.
My company recently hit the need for maps showing several of our own proprietary layers at various zoom layers, and considering how much of the data is already in GeoJSON, converting that to the Mapbox Vector Tile binary format (MVT) was a breeze to implement on our own servers, compared to rasterizing all of the layers and re-rendering whenever any part of our datatset changes.
I'm surprised about this as well. Even if they didn't want the overhead of running their own tile server, one of the best parts of the MVT format in my mind is that one-off tilesets like this layer can trivially be thrown into S3 and served extremely cheaply as well. That removes a ton of the effort they put into circumventing MapBox's limitations and is cheaper to host.
I'm a bit late to the discussion, but I'm glad to see so many comments here in response to one of the Utility API founders.
Just wanted to note something very important:
As a co-founder of a very new cleantech software startup, and as a guy coming into the energy industry initially with only a background in software, you need to understand that the energy industry (especially in the US) is incredibly diverse and is really an ecosystem in itself.
The market is not just regulated utilities and their customers. There are unregulated markets, power generators, ISOs, cleantech vendors, project developers, installers, financiers, insurance companies, and plenty more players in the space. Each of these actors serves a different role in the process of getting a solar array or battery up and running on a building or out in the field.
And to echo some other comments, the hardware is pretty much all there by now, with prices getting lower by the month, and the only thing that's missing is the right software to help scale deployment of capital, energy assets, installation labor, etc.
[Impending plug is coming]
In fact, this is exactly what my company, Station A (https://stationa.com), is doing. We've realized that in the commercial and industrial sectors, renewables are seeing slower penetration because the high soft costs associated with project development for the medium-to-small energy consumers. This stems primarily from a lack of access to critical data, biased sales tactics (solar developers only want to sell you solar, even if it won't make a difference), and the inability for project developers to really find the right locations to develop with the right technology in a scalable fashion.
This sector is ripe for disruption purely with software, as the hardware has become so much more commoditized.
Our current solution is to use a "toolchain" git repository that helps to automate installation of 3rd-party software (so that we're using similar versions of things like Docker, Python, etc.), but also creates a company-specific directory with installed scripts that can be added to the individual's PATH.
I have found the best way to get right angled USB connection that can be repeatably and reliably wired and is robust for rugged use is to create a small interface PCB in CAD software like Eagle and use a PCB mount receptacle rather than a straight inline/wire-to variant.
The PCB only needs to be big enough to land a through-hole usb connector [1] and then bring copper traces from each pin perpendicular to the connector to the edge of the PCB. You can even make the PCB quite long If you want to you as this will give you long copper traces you can expose to land the wire ends on, which is very useful if you don't have a great soldering iron or don't have much experience soldering. You could even 3D print an overmould to make it look professional, otherwise epoxy. Using this method you can get an right angled usb that's a tighter angle than anything you'll find online because it doesn't rely on cable bend radius to make the angle.
That's a pretty cool idea. However, I see two rows of pins on that connector, doesn't that matter? Are you making a two-sided PCB, or are you just soldering the pins together? I'm not sure if they're the same or not.
Credit goes to Hirose for the idea on USB Micro B, as they hint at it their ZX catalogue [1], but logically everything on USB devices is normally a receptacle_ so custom cables really is the only use for a PCB mount _plug_. If you download the catalogue and scroll to page 3 you can see example 3D renders on the second half of that page that might make it a bit more obvious what I was describing.
In actual fact for USB C you'd probably end up having to mount all connectors in the cables on tiny PCBs regardless, because of the potential requirements for signal conditioning and electronic marking ICs unique to USB C implementations. With so many pins on usb C without a doubt you'd need these tiny PCBs to be double sided to get all the traces to the PCB edge, and I wouldn't be surprised if some commercial usb C leads have PCBs with 4 layers.
I've only designed PCBs myself for the traditional USB PCB mount USB A plug (an Amphenol part) and Micro B plugs (Hirose ZX series parts) as these are very simple pinouts and are perfectly doable for a hobbyist I'd say if you want custom cabling. USB C looks significantly more complex and an in-cable PCB interface is the only way to go for that level of complexity.
I wouldn't play with USB Type C unless you know that you will never use them in any USB high power scenarios.
100W isn't the place for questionable soldering and thin wires.
I think the ultimate issue specifically with Facebook is not that you shouldn't give people voices online, but rather that when the voices form small, tight-knit echo chambers with zero outside challenge, the outcome can be disastrous. I believe this is why the author calls it a "walled wonderland".
