|
Top-level
 @bdon @openstreetmap True, I also suggest a parallel better managed structure here because of this: https://t.me/foss4g_ru/1463 I frankly have no idea how to answer, _why_ we need to change OSMF. It works perfectly as it is, the map is getting better, people are still making things, and we have Overture for a new facade. 16 comments
 @zverik @openstreetmap are salaried developers/mappers the best way to support the project? WMF and SO have salaried developers but not salaried contributors. Those are both vertically integrated - they produce a product that is directly consumable. That productization is key for visibility and then fundraising. OSM is less like wikipedia/SO and a lot like Linux: it is essential for so many applications, but insufficient alone. Most kernel devs are on member corp salaries (i think) @bdon @openstreetmap Support doesn't have to be monetary :) Currently there isn't _any_ form of support, but a degree of opposition to anything. It takes a certain frame of mind to participate in OSM, especially in Europe. @zverik @openstreetmap the malaise over OSM is the alienation of mapping labor - all the productization and thus value extraction happens outside of OSM proper. So commercial downstreams keep their “distributions” of the kernel proprietary with the minimum possible collaboration inside the OSM bubble. Mappers are turned off by potential exploitation so become suspicious of anything other than mapping for mapping's sake. A vicious cycle! @bdon @openstreetmap Looks like the question is allocation of resources! Corporations focus on preparing the data to deliver to their customers. Perfectly fine, that's what we're here for: mapping for people. It's developers we're competing after. OSM doesn't have good monitoring / validation tools, hence the downstream distributions. If wikipedia didn't filter lies, won't be there a commercial extract too? Corps have more resources, they get a better deal. OSMF doesn't, mappers get nothing. @zverik @bdon @openstreetmap This is not really similar to Linux at all then. Mapping and make data doesn't require the mapper to be a developer. Developers are needed for infrastructure. You also need expertise in design, and people with product vision who can lead and rally a team. It's hard for mappers to be developers, designers, product owners, all in one person. Many OSMers don't care about usability. Or if somebody actually sees their work. Many like me mostly use Google Maps and Waze. @cbed @bdon @openstreetmap I'll answer to each tweet :) So, this is indeed unlike Linux, but a bit like Wikipedia. They have got millions of editors, but very few developers — some volunteer, some they had to hire, because people rarely align to tasks. Another example is HOT. They are primarily mappers and community builders, but they have to hire developers to make their operations effective. An opposite example is the TomTom OSM team. They struggle with what tools other people made. @zverik @bdon @openstreetmap I am curious if you can expand more on the state of mind. I might have the state of mind, or I might be a pretty active mapper in Europe (for a significant part of 2016-2023 part time to full time) who went under the radar Swiss community is quite exception in my opinion though, and may be different than others, I really enjoy their mailing list @cbed @bdon @openstreetmap Well, in Estonia where I map (and in both other Baltic countries), there are 1-2 people who are mapping a lot, and also checking every other changeset for mistakes. And sometimes attacking in comments even smallest changes. I can see good intentions underneath, but also it's like the worst school possible, where you get punished for slight mistakes, but get no feedback on anything else. @zverik @bdon @openstreetmap Ah I see what you mean! What do you think are good positive Feedbacks? If we were Waze community, you would get badges but that's not much. In another world it's some crypto or cash, but that's not really what OSMers want (well I would accept $1/changeset of course). What are some ideas on more recognition? And is this also moving toward Gamification? @cbed @bdon @openstreetmap At SotM 2013 I proposed a "like" ("good job") button for changesets :) Small, but impactful thing. There could also be notifications like "XXX improved your shop" (but I see how it can hurt), more visible regional statistics, awards, (geo-)chat... Many things, games and wikimedia can provide inspiration. Definitely not monetary or physical incentives! @zverik @bdon @openstreetmap Yes! We thogihtt about this for Mapillary. We allowed commenting photos previously, like OSM in changesets. And we discussed a human feedback loop. Like Google Local Guides: how your data is used. Rob Atkinson at OGC has good idea on Provenance also, seeing how data is applied by who and where, as a way to know it's potential purpose too. I'd love Organic Maps to share the most viewed tiles or click count of POIs as a start. @bdon @openstreetmap The Linux analogy is good. Imagine if all but one developers are volunteer 20 years into Linux life. OSM doesn't have to pay directly. But they could pull their strings (that they don't have at the moment) to get people into corp. Or help preparing grant applications: I know of four OSM-related projects submitted to @NGIZero with zero help from anybody, all rejected. @zverik @bdon @openstreetmap Honestly I feel like your views align with Rapid a lot. The entire idea is to build an editor that supports mappers better. It doesn't force anybody to edit in a certain way, to use AI data, it just make things available and gives great UX, or aims to. @cbed @bdon @openstreetmap Frankly, Rapid is an import tool designed to circumvent Import Guidelines. It looks awesome, it's fast and great to use, you feel powerful with it. But the core function is large-scale imports. It primarily supports data consumers, not mappers. Like, the "Rapid" button is still the centerpoint of the UI. And I've written that in editors (and extracts), we're pretty much covered (though always not enough). Everything else suffers, e.g. changes management. @zverik @bdon @openstreetmap I think that is wrong. An import is an import. A user inspecting data to see if it is individually fit to add to the map is mapping/digitising. I do the same with EveryDoor when I look at a shop and add it to the map, never getting very precise (every fire hydrant is probably 50cm off). The center of the UI is the map, the Rapid button is a tiny part. AI data is like a 2% part. This has been explicitly written about the new releases, and dropping of MapWithAI name. |
Gregory's Server
@bdon @openstreetmap I'd start with "support". I don't feel supported when I map, it's like stackoverflow in there. I don't feel supported when I develop things, except for people using my apps and thanking me in person. I don't feel supported when I document or propose things on wiki and around.
To me, that would be OSMF's primary goal. It's stackoverflow's future we should be afraid of, not wikipedia's.