@mwpdx once @algebre installed an exact copy instance of OpenStreetMap, with the same interface and accounts creation and API working.
It was working fine with JOSM and Vespucci but what's the pupose if we cannot share and merge our new geodata with the whole OSMworld ?
Let's say I downloaded MainStreetRoad to work on it offline and in the mean time `SomeonElse` (this is a real nickname of some DWG OSM guy by the way ^-^) is working on it via ID Editor then saved it. When I'll try to submit my own work, I'll have to resync because new changes occured on that street.
The problem is that sometimes, it is not easy to solve a changeset conflict.
If the toot is about infracost sharing ressource, I agree 100%.
@BoF Excellent points. Even if we could have an ActivityPub based platform for mapping and only talking upstream to OpenStreetMap via API... that would probably accomplish the same goals. But, then it's not a truly decantrilized solution because of a centralized API being part of the mix.
@marsxyz @mwpdx yeah well... more or less. It's ok to find a point with coordinates and get some rough idea, but it often doesn't know cities or places, and frankly I take its navigation more as a suggestion and keep my brain awake when following.
I say this with a lot of love, I'm contributing to OSM too and I hope it becomes a standard in a few years. And anyway, asking for indications makes for good social activity
Over the years navigation guidance has gradually switched from suggestive to directive. Most of the local ride-sharing drivers, for example, have no idea what the best route is, if they don't use their smartphones - while the local taxi drivers will know exactly (or close enough) where you want to go.
Have we not have enough news of people driving off road (or worse) because they followed their SatNav instructions rather than using their common sense?
There has been /some/ work on this already: https://wiki.openstreetmap.org/wiki/Portable_OpenStreetMap
But I haven't looked into them too much,
@rory @mwpdx An interesting aspect I didn´t think of.
Offline map use, edit and later merge is decentralization? But it would be a valuable option.
“Go Map!!”, “StreetComplete”, “Kort Game” and POI editors are steps to offline editing. Walking Papers as an App, yes!
There are many offline map-using tools. I use Pocket Earth but have some wishes: more style options; and editing, include “GO MAP!” If I would have a second live, I would try to code and improve something similar (as PWA).
I am now lost in options and projects. I would like to contribute code, vector clients prefared, may be adding 3D rendering to it.
It´s just my dream to
- Zoom in a vector map, unlimited!
- See any existig OSM data at last.
- Transit form the top down 2D view into a steet level 3D viev.
@karlos oh boy, this is something beyond my scope unfortunately. I believe pulling in tiles, or at least recreating them on the fly server-side as SVG based off of some kind of data source is doable. Maybe even utilizing HTML5 canvas could be an option... I dunno. This random thought of mine never factored in technical aspects.
@mwpdx 1/3: Communicaton:
#OpenStreetMap has its communication system, kind of web based Email, old-fashioned in my eyes.
I had the idea to start an OSM-Communicator/#Messenger service&clients(by PWA). Centralization is an advantage for such systems. But now I would prefer, the federated protocols. Image, all the million #osm contributors would become part of the federation!
To make osm use the federation, it must become better and comparable to all the commercial services like slack.
@mwpdx 1/2: Data base:
I thingk, the actual #OpenStreetMap data base is highly redundant and efficient. Decentralized is only the use of the data, renderer and tile server, Overpass.
Schould the editor server be decentral? Well, it could, because edits are (mostly) related to “small” areas.
And by the way, I miss open vector tile servers.
@mwpdx 1/3: Map use:
I am sad about #OpenStreetMap is not focused of using its data. That’s left to commercial mostly.
The main page map is not a service for map use but for data editing. That’s fine, and I would keep it this way.
What I miss is a map for users. And I would build it as client side renderer. That would offer wide flexibility because the user will define/refine the map style and content.
Follow friends and discover new ones. Publish anything you want: links, pictures, text, video. This server is run by the main developers of the Mastodon project. Everyone is welcome as long as you follow our code of conduct!