What I see this helping is the proliferation of OpenStreetMap applications. For instance, OsmAND and Maps.me (as I vaguely understand it? Please correct me if I'm wrong) have their own data files served from their own servers, copied from the source OpenStreetMap DB. Periodically, the app developers update the data on their servers, and users soonafter get those updates. (I like this system at least as an option; offline maps are convenient and at least in theory more private.)
Where it gets interesting is if you fork the program. Perhaps OsmAND and Maps.me aren't so fond of having their data used by forks of their applications, for whatever reason. This creates a pretty major barrier for people who want to make a fork; they now have to maintain a data server. (Fortunately for F-Droid users like me, there is a fork of Maps.me simply called Maps. It removes ads, so I imagine Maps.me isn't interested in subsidizing it in the form of server access. I'm not sure what they worked out as far as storage.)
It always occurred to me that they could mostly solve this problem with bittorrent. But if the entire database is torrent-like with Peermaps, that's even better. It could remove one of the barriers to experimenting with new apps. Or so it seems to me.
Yes the OSM data format is good for editing, but includes everything, and isn't sorted or anything like that. So data consumers process the data, keep what they need, clean up data, process it and make into whatever format they want/need.
You're right that the 'whole planet' is a lot of data to process, and this can be a barrier. But there are regional extracts of OSM data available (e.g. http://download.geofabrik.de/ ). So one can always maintain just data for your region
It sounds like you're saying that the reason these applications have their own data files is that they tailor it for the application; nothing to do with bandwidth issues.
In that case, I don't see how peermaps would make things any different for them.
I think "the peer web" and "decentralized web" / "dweb" are emerging terms that are trying to describe all of the efforts to make software work peer-to-peer without centralized servers.
I am referring to https://peerweb.com, an in-development project of mine. There are no public-facing docs at the moment, but I will be providing updates on my Twitter.
This type of thing seems very useful if you are ever going to get a viable public alternative to Uber as an open network and platform that providers can build off of.
I wonder if peermaps makes it convenient for peers to update maps? Because OpenStreetMap is missing a lot of key information such as not having addresses in many places.
But a peer to peer mapping technology with a more complete OSM plus a scalable cryptocurrency such as Eth 2.0 and maybe a smart contract plus some distributed coordination for bidding on rides could make things more practical.
Right now arcade.city is doing something similar but is limited by some of the factors I mentioned.
What would you do differently about OSM contribution? Presumably peermaps pulls from the main OSM database. Would a distributed system somehow make contributions easier?
Where it gets interesting is if you fork the program. Perhaps OsmAND and Maps.me aren't so fond of having their data used by forks of their applications, for whatever reason. This creates a pretty major barrier for people who want to make a fork; they now have to maintain a data server. (Fortunately for F-Droid users like me, there is a fork of Maps.me simply called Maps. It removes ads, so I imagine Maps.me isn't interested in subsidizing it in the form of server access. I'm not sure what they worked out as far as storage.)
It always occurred to me that they could mostly solve this problem with bittorrent. But if the entire database is torrent-like with Peermaps, that's even better. It could remove one of the barriers to experimenting with new apps. Or so it seems to me.