 What I think happened when @linuxfoundation took Overture Maps: Companies and foundations want to support OpenStreetMap. They see the huge community and the impact we make to the world. Foundations were made to support projects like ours. But. We already have a foundation. OSM Foundation. They rightfully expect OSMF to support project grows, that is, to support mappers and communitiy and tools for editing and validation. Tools that would make Niantic stuff easily identifiable and revertable. 60 comments
You might think, this is normal, OpenStreetMap always relied on people, and companies rarely did anything good for us. Which is wrong, in OSM history years like this were rare, maybe 20%, mostly at the beginning. That thing that's happening everywhere, with mass layoffs and axed projects, reached OSM indirectly. We're on our own now. So what the community (well, @simon yesterday) is saying, we see Overture, Linux Foundation, and Niantic moving big money, why they don't support the commons, us? And my points are: 1) €30k for participating in OSMF won't do anybody good, and everyone sees that. Niantic were smart to not subscribe. 2) Overture Maps delivers. They are worth their money, and they will make the world use OSM. They are good guys. 3) Supporting mappers is like supporting openssl. Nobody even thinks of how the map is made until something breaks. And OM protects people from obvious breakages. 4) There is no way to support mapping currently. At all. Not Niantic's or LF fault.  @zverik Good thread, you’re basically on target. Overture is an outcome of OSMF’s traditional policy to remain as small as possible, plus a decade of the originating companies repeating each others’ work on OSM mapping and validation and finally realizing they have a common interest.  @ian @zverik That lost decade was a key ingredient: those of us who’ve moved through these companies working on OSM-adjacent things kept hearing the same stuff in our informal chats over the years, like “oh yeah they’re trying to figure that out at the fruit company too but nobody’s allowed to share notes unless we create an exec-approved structure to do it in” … this is what LF is for.  @zverik You are really missing the point. While it would have been nice if Niantic had supported the organisation producing what one of its major products is based on, that isn't a requirement and nobody suggested that. What is despicable on any count, is refusing any communication while its customers vandalize #OpenStreetMap. For example communication on how to best mitigate said vandalism. @simon If OSM were successful, Niantic would be one in a thousand similarly influential map users. To me, it's the same as the current e-scooters on sidewalks debate: the problem is not in scooter drivers or road code, but in the infrastructure. OSM infrastructure hasn't significantly changed in 12 years, and that's not a LF's fault. But, on the other hand, we'll have vector tiles any day now. (okay I'm missing the Fastly and Overpass things that happened in 2021, but still, that's not about mapping) @zverik I would suggest sticking with the facts including not claiming that #OpenStreetMap has failed. OSM infrastructure is day and night different than it was in 2012. Not just technically but organisational too. Heck back in 2012 the most popular editor was flash based Potlatch 2. Current contributors wouldn't even know what that is if you hit them over the head with it. @simon Good thing Mapbox was there for us back then! I'm not saying OSM got worse or anything. Ten years ago I wouldn't write all this because nobody expected anything from the Board or OSMF. The governance did improve a lot with and after Allan. It's just their strategy is still "make an app idk", and I want to affect it — but I don't have the hours to devote to working on the Board. Alas in this I feel I'm turning into Christoph :( @zverik actually Mapbox back then was a good example of how a symbiotic relationship with a commercial entity can work to the benefit of both sides. Mapbox repeated what they had already done with Tilemill to kickstart the company enough that they could really get going. It was clear that this wouldn't fly without buy in from the OSMF and OSM at large, and they talked to the OSMF and once the grant was awarded the work on both the website and iD was done not against, but with the community. @simon @zverik The Mapbox staff doing the work at the time described how difficult it was to work with the community: enduring abuse, devoting time for psychological recovery, limiting time spent on the project, and swiftly exiting it once minimum deliverables were complete. Source: talks from team members ca. 2014, not recorded. Great software came out of it, but also a general understanding that the OSMF can’t support this kind of work.  @grischard @migurski @simon If that's not the case, I'm happy to be found wrong! @defuneste @simon not successful ≠ failed. I meant it in terms of adoption as it is, a universal public geodatabase, not just a background layer or anything. So I'd say that's a failure of most companies to see the value in OSM, not of OSM itself. Even here I would separate "not seeing value" and "not seeing value to provide some support". Seeing the list of company that use OSM, and pay mappers it does not seem they do not see value. Even as a "public geodatabase" it is quite good and way above some corporate mess (yes I works in private sector and I have see some nightmare). I should rephrase that: "data does not bring value, product does" (say a lot of folks) @zverik @sotm our 2024 budget https://osmfoundation.org/w/images/0/06/OSMF_2024_Budget_v2024-02-23.pdf includes 12k£ for local chapters and 133k£ on engineering, which includes editor software. This is a lot more than previous years, and shows how supporting mappers is a priority for us. Our income is our limit. Afaik, Overture and Niantic haven't spent a penny on supporting mappers. @grischard That's great, I know of this and I support this! I just wish that a) OSMF had more income (which doesn't appear magically, it's daily work setting up partnerships and planning, and that's not a job for volunteers), and b) it did more direct outreach (like microgrants and pizza funds and community funds), in addition to €80k for iD development (essential) and like €24k sponsorship between EWG and LCCWG. @zverik tell Niantic to actually become a member then :). There is a microgrant budget for both of these - again, limited by income. I hope you have cool ideas for a microgrant in the pipeline. @grischard One of the points I mentioned in my rant, I don't believe an extra €30k would do any change. This model would work if we had 100+ organizations as platinum members. With just two (one of which did not intend to), it is unsustainable and indicates something's wrong with the entire corporate sponsorship framework. I'd prefer for big companies to wait until OSMF comes with a program aimed at big donors that demonstrates their money would indeed make the map better. @zverik @grischard €30k is the membership cost, not the upper bound for how much a company can contribute in a given year if OSMF has a particular funding drive / specific program that might be of interest to them.  @zverik FWIW, open source foundations talk amongst themselves all the time, including on a mailing list that has been active for over a decade. OSMF has, for various reasons (some good, some bad), never really actively participated in that conversation (either online or IRL). (This is not to defend LF, who are not always great partners either, though for different reasons.) @zverik I think what happened is that companies and foundations wanted some sort of "handle" on OSM and didn't understand OSM enough so they had to stuff it into a box they could understand. A box that is not a movement but a proper corporate entity they can deal with. Something with a big budget, lots of employees, and a leader you can put pressure on. It is possible that the OSMF could have pre-empted that by becoming such a corporate entity. It would have been a capitulation. @woodpeck No. Enough. I have read the same sentiment since I joined OSM, and at first I agreed. 14 years later I see nothing has changed, and mostly because you are influential in OSM and this all looks logical. But what you are saying is, OSMF should remain powerless and amorphous, so that loud voices kept weighting more than an organizational strategy. Becoming a corporate entity that has a roadmap and is easy to talk to is not a capitulation. It's growth. @IvanSanchez @woodpeck hah I knew this word would doom the sentiment :) The goal of @openstreetmap is indeed growth. More mappers, more map data, more tools, more users. There are nuances, but I'm sure if you ask around, 99 out of 100 times you'll get this. For OSMF, the goal is supporting the project as an organization. Meaning, it should grow with OSM, not fall behind. What I meant, was not numeric, but organizational. OSMF is still dozen IT guys in a trenchcoat, not an entity to talk to. @zverik @IvanSanchez @woodpeck @openstreetmap I am sorry but to me, OSMF is to support the environment for OpenStreetMap.  @Melaskia @zverik @IvanSanchez this is why I love apps like @streetcomplete and @everydoor. They're highlighting another level of #OpenStreetMap and making it much easier for people (incl me) to add that data. There are many types of data, where we are far from complete @amapanda @Melaskia @zverik @IvanSanchez @luis_in_brief Since I've had anything to do with OSM, a good 15 years, enabling low barriers to entry has been the dominating mantra, and there have been wide variety of apps doing exactly that. They do not drive long term engagement any better than the web based browser iD, and actually most of them are so much worse at doing that, that we are not really doing OSM a service by directing people to them. ... @amapanda @Melaskia @zverik @IvanSanchez @luis_in_brief ... Note on the side compared to WP even iD makes contributing trivial so we are not talking about a high bar to entry to start with. I had a longer discussion on a related topic with @zverik here https://en.osm.town/@simon/112206877070048055 @amapanda @Melaskia @zverik @IvanSanchez @luis_in_brief My hypothesis is that making things easier also reduces the value of leaning how to contribute and makes the whole thing less of a rewarding pastime. There are naturally similar scenarios in all areas of human endeavour. @amapanda @Melaskia @zverik @IvanSanchez @luis_in_brief Back on the maintenance topic. There has been this concept that this would be done by the drive by contributors that the "simple" tools are targeted at, and while that can't be completely dismissed, just the volume and nature of what has to be maintained would suggest that that is not going to happen. Just consider that it isn't really #OpenStreetMap these days ... @amapanda @Melaskia @zverik @IvanSanchez @luis_in_brief .. but more "OpenBuildingMap" keeping that up to date, regardless of (insert you fav buzz word here) tooling is simply uninteresting drudgery that is only going to be done by people that are actually committed to the cause. ... @amapanda @Melaskia @zverik @IvanSanchez @luis_in_brief ... Factor in the difference in contribution volume between incidental and even just average regular contributors (easily 10^5) and I believe you can't come to a different conclusion that what we really need is more highly engaged and dedicated contributors. @simon @amapanda @Melaskia @IvanSanchez @luis_in_brief Yeah I came to a similar conclusion. Not dismissing drive-by contributors from OMaps and SC — they are essential because some map is better than no map. But I really don't know how to find highly engaged contributors. How to find people that look at the map and think, oh, okay, this might be the thing to which I will devote every day of the rest of my life. All I can do is make tools to ease and speed up our work, so we focus on fun bits. @simon @amapanda @Melaskia @IvanSanchez @luis_in_brief That is actually an excellent point! Because "contributing to OSM" is a non-goal, much like "reading a book" or "caring for the environment". A goal needs to be achievable (e.g. medals in StreetComplete) and social (so that a person gets some praise). Missing Maps events got that quite right. For me, OSM is kinda like pokemon: I like to collect everything. That's why @everydoor is like this: clunky but super effective at a scale.  @zverik @woodpeck @openstreetmap I disagree that growth is a/the goal of OSM(F). The *raison d'être* of OSM was the lack of availability of geodata, and growth **was** a means of achieving critical mass for affecting policy, which IMO is the goal. Drive a fear of becoming irrelevant into mapping agencies. I'm totally biased because I'm a Spaniard, and we've got creative-commons-licensed 10cm imagery and 5k vector maps and gazetteers and whatnot. Mission accomplished. We can chill and have fun. @IvanSanchez @zverik @woodpeck @openstreetmap That’s the European perspective on OSM in a nutshell: works great, what’s all the fuss? The corporate perspective when I was involved looked at the places where OSM was not yet successful: for FB, mapping outside rich countries was important. For other Overture partners, consistency and depth of data for navigation & wayfinding were needed. Since that time TomTom has built an entire data strategy around interop with OSM: https://www.tomtom.com/tomtom-orbis-maps/ @migurski @zverik @woodpeck @openstreetmap I agree with you in the sense that I acknowledge that US-based corps would like OSM to cover their wants/needs. But you're kinda making @woodpeck 's argument here: corps struggle to grasp this grassroots thing made by a handful of british guys and a dozen german hackers in a trenchcoat. Corps will want to sculpt that into something they can have some control and extract value out of. @migurski @zverik @woodpeck @openstreetmap Which leads to the *ethos* of the issue at hand: What do we want OSM(F) to be; what should OSM(F) be? It can be a crowdsourcing tool so that US-based corps can extract value out of it. Or it can be a GIS playground so that anarchist hackers can play around and invent new things. I do not think it can be those two things at once. And I'll choose anarchist hackers over US corps, any day. @zverik @linuxfoundation quoting myself: "It needs to be pointed out that if the OMF founders were actually doing this for the good of humanity, they could have simply open sourced their validation tool chains and financed the OSMF running them. But they didn’t." @simon @linuxfoundation That's... wrong on so many levels. OSMF is not equipped for having developers of this scale, even managing the workforce, managing the infrastructure needed for big data processing on that scale, and many other things OMF requires to advance. It would require OSMF to shrink OSM to like 5% of its focus. @zverik @linuxfoundation why would opensourcing the validation tools stop devs from Meta & co working on them? @simon @linuxfoundation OSM is not about open source, so this requirement is pretty random. Given that virtually nobody in OSM would be able to understand or run it, so it might benefit only other corporations with a similar stack — that imo would make it worse, not better, for OSM. |
Gregory's Server
Nobody in OSMF asked to be part of LF. And we shouldn't. Foundations usually don't talk among themselves (I guess?)
So there is another group. For packaging data for consumers and promoting open geodata to businesses. On a scale never seen before. They needed a foundation, and LF stepped in.
Why wouldn't they? It open data. Participating companies have a good track record in supporting open source. Bringing open data to more people — what's not to like? OM competes with Google, not OSM.