Views from the OpenStreetMap Foundation on the launch of Overture

mikelmaron · 2022-12-22T16:33:49.224Z

The OSMF Board has released a statement on Overture Maps. Welcome the community’s feedback, questions and discussions

https://blog.openstreetmap.org/2022/12/22/views-from-the-openstreetmap-foundation-on-the-launch-of-overture/

ZeLonewolf · 2022-12-22T16:40:34.788Z

Thanks Mikel and the OSMF board for this measured, reasonable, and professional response and outlook. I believe there is a place in the OSM community for hobbyists like myself, people that work with maps for a living, as well as companies large and small, to follow their interests and solve their challenges with the tools available to them. I appreciate that the board is engaging and looking out for the project’s best interests.

RobJN · 2022-12-22T16:50:31.811Z

What about OSMF becoming a member of Overture Maps Foundation? Would be easier than being on the outside and hoping they provide regular updates.

TomH · 2022-12-22T17:05:29.003Z

Well we’d have to spend $300,000 a year to actually get any sort of say in anything - anything less than that is basically just an observer status.

RobJN · 2022-12-22T17:27:15.928Z

And observer status is better than non-observer status!!

mikelmaron · 2022-12-22T17:31:07.803Z

Anything is on the table for discussion for best framework to engage, whether under currently shared tiers or in some new arrangement. Likewise, could consider Overture becoming corporate member of OSMF. It’s too early to say what’s the right approach.

RobJN · 2022-12-22T17:34:28.803Z

Sorry I thought you were asking for community feedback . Joking aside my feedback is that you should put some formal arrangement in place rather than rely on ad hoc updates from each other. Both becoming members of the other feels like a good place to start.

ezekielf · 2022-12-22T17:56:46.111Z

Great post, Mikel. Thank you. I am cautiously hopeful about this new endeavor, despite my initially skeptical gut reaction. I appreciate the board approaching this with openness as well as caution. As for the possibility of OSMF becoming a paying member of Overture, I’d wager the value of the OpenStreetMap worldwide dataset is far greater than $300,000 a year. Just saying .

mmd · 2022-12-22T19:41:08.517Z

One of the four Overture pillars “Collaborative Map Building” specifically mentions “civic organizations” and “open data sources” as one of its sources.

Do you think this could undermine our efforts to convince these organizations to release their data under ODbL-compatible clauses? How can we ensure that open data will continue to be available to OSM in the future?

Mateusz_Konieczny · 2022-12-22T19:48:22.593Z

RobJN:

And observer status is better than non-observer status!!

It also suggests that OSMF supports/welcomes their activity - and doing this may be premature as it is not entirely clear what are their plans.

RobJN · 2022-12-22T19:56:14.403Z

This is true and it did cross my mind. Personally I don’t think it’s something that cannot be easily managed (e.g. through a clear statement, not being listed on the site as a member, etc). In my mind the advantages of being in the loop outweigh the disadvantages of being seen to endorse the project.

SimonPoole · 2022-12-22T21:52:17.931Z

To actually have a say you are missing a 0 in that number.

fititnt · 2022-12-23T03:24:09.339Z

From the link:

The technical problems that Overture is addressing, such as quality checks, data integration, and alignment to schemas, are valuable for any map data provider. We know data consumers have been working on these challenges in isolation for too long, and by bringing these issues out into the commons, and open sourcing the tools, we can all benefit.

I would ask OSMF board members to read propaganda announcements from OMF with a grain of salt, otherwise your nexts public announcements will be used by OMF to devalue volunteer contributors on OpenStreetMap in one of the worst moves in the past decade that could sideline OSMF. Despite multi millionaire investment to have voted on OMF and the requerimento of 20 engineer’s, to my knowledge they’re likely to be less efficient than any existing OpenStreetMap organization could do by simply organizing itself and do post-processing of its own data. [1]

Some early evidence for it are @SomeoneElse post or prior analysis and other topics on the founding members actually likely to implement what they’re testing over last years isolated before join up together. I’m actually quite deceptioned if they’re “grand triumph” over OSM data exchange encoding is just a variant of Apple’s Indoor Mapping Data Format (which by itself have vocabulary heavily based on OSM Wiki, I will avoid say even more here (because I’m quite busy), but by the time OMF starts it’s first public release next, if you OSMF decides to side with your existing developers to make a schema-stricter version of OSM data (without any groundbreaking advancement, just to organize it) you will deliver better quality over OMF because their implementation is inferior.

mmd:

Do you think this could undermine our efforts to convince these organizations to release their data under ODbL-compatible clauses?

Both the license and the branding OpenStreetMap™ not be forgotten by big corporation’s is a clear reason to me, even if just in Brazil alone, to do whatever it takes to not allow OMF to take the role of OSM/OSMF with government and civic organization here. So, if OSMF decides to not have its own schema-stricter version and let a foregueiner foundation do it while spreading FUD about OpenStreetMap data, then we here will do at least for our region.

RobJN:

What about OSMF becoming a member of Overture Maps Foundation? Would be easier than being on the outside and hoping they provide regular updates.

As even the @cRaIgalLAn (while as candidate) on his comments “The OSMF Board members are bound by British law to owe a fiduciary duty to OSMF” (link to wiki) let me remember again here that the OSMF board members cannot do something knowing to be harmful to OSMF and (even if any board member have no conflict of interest on companies behind OMF) being member of another foundation that uses OSM data and, by public intent, already signaled to go in same areas OSMF would go, it’s a clear risk for OSMF existence, so even before ask public opinions of OSM collaborators, OSMF would need contact lawyers if you can be a member at all of OMF even if they offer it for free.

Mateusz_Konieczny:

It also suggests that OSMF supports/welcomes their activity - and doing this may be premature as it is not entirely clear what are their plans.

Yes, I would agree with that. Even if they somewhat consider OSMF viable as any kind of membership (and, again, this could conflict with their clear goals) the like-hood of OMF take decisions that could conflict with OSMF is high.

However, might be simpler OSMF deal with near same corporate members of OMF (which I suppose except by AWS, already is the case). Assuming the context that both foundations can conflict with each other, OSMF could be a better alternative.

[1] Add to this that so called deep leaning they advertise to clean data have serious limitations over handcrafted rule-based expert systems for structured content (such as tagging, and even geometry checks), and the advancements on road detection from high quality imagery is one of the few cases deep leaning is a great implementations over other strategies on artificial intelligence.

stevea · 2022-12-23T03:53:07.643Z

Agreed. I don’t think it’s a secret that “the best maps in the world, of the world, for the world” is a highly valuable asset, one which “single membership” in has annual “costs” (to influence, to guide, to shape, to steer…) of “low millions” or “seven-figure amounts.” Quite likely, even much more than that.

OSM does this currently because of the dedicated effort of volunteers (somewhere between hundreds of thousands and millions). Even if each of us considered that an entire year’s worth of effort is “worth a few dollars,” we’d be way, way, underestimating what “real costs” are of what we volunteers contribute to OSM each year.

In reality, what seems to be the “uncrackable nut” (for any one human, even as smart as Steve Coast is, he is “only one human” who had a terrific idea) is “how do we GET the ‘best map on Earth’”? Let alone pay for it? As crowdsourcing (OSM for nearly two decades) has proven “pretty darn good, though with its issues…” and AI / ML might offer an edge in the future as the tech is real enough to be making a daily (positive? the jury seems to still be out) difference. Humans are evolving, along with our tech, along with our social tinkering, along with our telecommunications and (software, big data…) infrastructure to produce these things, yet the evolution is “only underway,” it is not (yet) completed, or even beyond what might be called “a rough and ready, working 1.0.” (Some feel OSM isn’t even that, calling it “an 0.6 dinosaur”).

Overture, on its surface, seems like it is complementary to OSM, or “so they say.” I think what we have now and will going forward as this emerges / releases is “a horse race,” where it will be exciting, and there will be those who pull ahead early, those who catch up and pass in the backstretch and who knows how it will end or “who will win?”

Think of this in the broader context, everyone. And if you’re in OSM (you are if you’re reading this, quite likely), you have a part to play: both IN OSM, and as part of an observer of and perhaps even participant in “complementary” efforts.

dieterdreist · 2022-12-23T10:22:56.000Z

As crowdsourcing (OSM for nearly two decades) has proven “pretty darn good, though with its issues…” and AI / ML might offer an edge in the future as the tech is real enough to be making a daily (positive? the jury seems to still be out) difference.

what has not been talked about is AI writing the tag documentation. There are thousands of undocumented tags and only few people writing documentation, this is a field where it could be very interesting to see how AI could leverage to its full potential. Also AI generated presets would leverage the influence of ML to unprecedented scale. There are only so many roads in the world to be discovered on aerial imagery, but tags and presets have no limits. Then we might also need AI to write the rendering rules though

stevea · 2022-12-23T10:32:50.057Z

“AI writing tag documentation,” indeed. That’s one of a number of (apparently needed) possibilities.

In the several hours I wrote what I wrote, I’ve been thinking more about this. What I’ve come up with is something like this: let’s say that in 5 years, I see how Overture makes real, positive contributions to “better map data” or “a better map / mapping experience for all” in ways quite similar to how OSM does what we do. And let’s say that the effort invested is “super-powerful,” like “quite leveraged.” Let’s say that instead of me spending my usual 100 time units per week (could be any amount of time) on OSM, I spend 99 on OSM, but I use that extra 1 to contribute to Overture. But as a result of that, “somehow” (because of Overture’s “powerful leverage” potential, “small effort, big bang for it”), “people’s mapping experience” isn’t now valued at 100 higher because of my weekly efforts, it is now valued at 150. As in “wow, by using Overture, I get many times more value for my efforts.”

I might choose to invest in that. Or, I might not. And my OSM friend over there…what about her? Well, she can choose to invest in Overture, or not, also. We’ll all “vote with our feet.” It won’t (necessarily) mean OSM goes away, more like “OSM has a force multiplier.”

I don’t KNOW that’s what’s going to happen with Overture. But it could. And I will be keeping my eyes open as to “how complementary” it actually becomes. And I’ll act accordingly, even knowing that “things can, do and will change in the future.” But it’s all too early right now for such specifics. So, I’m glad we’re discussing it.

dieterdreist · 2022-12-23T12:10:07.151Z

stevea:

And let’s say that the effort invested is “super-powerful,” like “quite leveraged.” Let’s say that instead of me spending my usual 100 time units per week (could be any amount of time) on OSM, I spend 99 on OSM, but I use that extra 1 to contribute to Overture. But as a result of that, “somehow” (because of Overture’s “powerful leverage” potential, “small effort, big bang for it”), “people’s mapping experience” isn’t now valued at 100 higher because of my weekly efforts, it is now valued at 150. As in “wow, by using Overture, I get many times more value for my efforts.”

I might choose to invest in that.

but wouldn’t the consequence of this not be investing 100 time units on Overture, to make a 5100% contribution? Your example sounds nice if you spend 1% on Overture, but it also illustrates the menace this could pose to OSM.

stevea · 2022-12-23T19:52:35.947Z

I certainly see your point, as it is valid, though “taken to an extreme.” Though I think that my point includes the perspective that “balance is what is important.” There is a natural feedback loop with “balance” such that, over time, the “right balance” can be found. A distinct potential danger is that AI / ML yield results that can be so quick, that the stage during “getting the balance just right” can quickly skid out of “easily manageable,” and happen too quickly, so much so that there isn’t (human, thoughtful) opportunity to “correct the situation.”

Plus, I don’t see it as likely (it’s possible, yes) that Overture would completely overtake OSM. So putting my efforts 100% into it (instead of OSM) probably won’t happen, as it wouldn’t make sense for it to happen. A “sane balance,” seen as such by sage OSM folks? Yeah, that’s possible.

Again, this is all speculative right now, it’s too early. And I don’t know if Overture will be something I can participate in or not: it might be a “billionaire’s black box” and I can’t touch it. That does concern me.

SimonPoole · 2022-12-25T12:32:05.730Z

fititnt:

I would ask OSMF board members to read propaganda announcements from OMF with a grain of salt

I don’t think you should read too much humble pie eating in to the euphemistic diplomatic response by the OSMF. Between the lines it says ‘we’re pissed off that you didn’t talk with us beforehand and we expect you to behave better in the future’, maybe it’s a bit too subtle for an international audience.

Mateusz_Konieczny · 2022-12-27T13:00:37.962Z

To be more specific, note for example

While many details of Overture remain unclear, the OpenStreetMap Foundation is interested to get a better understanding of the project

many questions remain. We have encouraged Overture to engage with our community and to share their plans publicly.

This stated commitment

explore the best ways to facilitate ongoing communication, decision-making, collaboration, and support

fititnt:

I would ask OSMF board members to read propaganda announcements from OMF with a grain of salt

Maybe there should be also translation to “international English as a second language, without subtlety”