So, this is the first public repository on the OvertureMaps organization.
here the direct link.
Edit: the repo was created 2 weeks ago.
The link to the GitHub repo is dead and I canât find it through Google search. So maybe they took it down or something?
Ok, I think the initial repository was just a test, the real one is live now (likely not just formaly announced).
Live page:
5 Likes
3 Likes
https://docs.overturemaps.org/gers/scenarios:
What if the company has proprietary (better) road segment data extracted from their sensor network that they do not / cannot share?
âHowever, by sharing just the new road segment geometries with Overture (not the entire proprietary traffic feed), Overture will add it to the Overture corpus and generate GERS IDs accordingly. The Overture road network will be improved and 100% of the companyâs data feed can become GERS-enabled.â
Would those âimprovementsâ then be fed back into e.g. OSM?
The OSM community has a basic quality expectation, but if I understand correctly, there are legal possibilities for this, in my opinion.
see:
- [overture faq] âOverture data will be available for use by the OpenStreetMap community under compatible open data licenses. Overture members are encouraged to contribute to OSM directly.â
I assume that would require citing them as the source for the contribution, which seems like a good avenue to potentially provide them with cheap advertising and a way to easily siphon off users. If thatâs the case it probably wonât be as beneficial to OpenStreetMap as it might seem at first glance. Especially in the long arch of time where 5 or 10 years from now there will likely be hundred of thousands, if not millions, of references to Overture Maps associated with objects. Rather parasitic if you ask me. Although again thatâs contingent on users citing them as the source for the contribution, but I donât see why they wouldnât.
1 Like
I thing this question should be more framed as âCouldâ not âWouldâ. I know of no indication that the OMF intends to actively contribute to OSM.
In practical terms as an OSM contributor you would probably want to use any such (currently purely hypothetical) data more for QA in any case.
3 Likes
No need to worry about that as they donât need it - just look at the names associated with Overture!
5 Likes
46 posts were merged into an existing topic: Overture Maps first dataset release
Everybody realizes, of course, that Overturemaps.org is watching every word of this discussion. âTheyâ and âtheir AIâ are parsing every word as hard as it/they can.
I suppose at some point we have to be real and say âand after that, lawyers sort it out.â Maybe not.
This is a âgrand conflictâ being deliberately constructed, if nothing else. OSM, we do well to deflect against too much time-wasting by deliberate time-wasters (like Overture?). A long-term strategy might be to think (hard) about how much âOverture is deliberately trying to spoof usâ might be true. A mild effort on offense can become (has become?) a serious effort on defense (by OSM).
The trick for OSM/us is to âhose off into a bucket of noiseâ what is exactly that, versus how much is an actual threat. Tricky balancing point, that.
Overture is âput in a box.â OSM/we continue to draw boundaries around it (versus vice versa). Eyes open. Yet, âyou are only this smallâ (and our fingers are seen to be squeezing a very small distance).
I donât want to dismiss a real threat, if it is one. But, hey. If noise isnât noise, and we call it out as noiseâŠ
3 Likes
To ask a delicate question: is there anything in the OSM ecosystem or community that the promoters of Overture find offputting? Is any of that issues that we could do better on for our own sake?
Iâm purely speculating here: I believe Overture finds OSM âoff-puttingâ because Overture cannot own / control OSM as it wishes. It seems to bother Overture greatly that a successful crowdsourced project full of altruism, damn good data, a real (kind, helpful-to-othersâŠ) community-ecosystem which continues to grow over years (decades, really) into not only some of the best mapping on Earth, but a SYSTEM of mapping, all based on volunteers giving of ourselves and our local (and wider) geographical knowledge. That rocks.
Iâm not saying that with aggressive lawyers and/or billions of dollars, âsomething elseâ canât âcompeteâ with OSM â it probably could or can â but Alphabetâs Google Maps, while it might be considered âa competitorâ to OSM, still has many shortcomings compared to much of OSM. Same for TomTom and Apple and Bing and others that are out there. I donât want to ârest on our laurelsâ and smugly insist âweâre the best damn map on the planet and nobody can touch usâŠâ because with that sort of hubris, we are easily dislodged from our august and likely premier position. OSM must always be looking in our rear-view mirror at âthe approaching competition.â
Perhaps one of the things that OSM might âdo better for our own sakeâ is to better articulate âwhat our place is in the world of mapping platforms.â Iâd like to say we are one of the premier mapping platforms, with some of the best data, constantly improving (as are our tools, community, documentation, education, outreach, ability to help newer users with guidance and improvementâŠ). But if thatâs not true, and âsomebody is eating our lunch,â well, letâs roll up our sleeves and get busy being better.
It can be difficult to talk about such things, as such âquantificationâ (such as saying âweâre the best mapâ) isnât easy to do and especially because doing so can be highly subjective. But in the cold light of day, OSM is able to look at something like Overtureâs recent .alpha-0 release and say âkinda junky, with old, noisy data from the trash heaps of commercial social media, redolent with the low quality one might predict comes from such an endeavor.â I donât think thatâs too far off the mark, but of course, it behooves us to keep our eyes wide open (towards additional, future releases).
The first pickle out of the barrel often is funky. What matters (SHOULD matter to OSM) is how much better Overture gets in the future.
5 Likes
The fallacy here is assuming that quality has anything to do with success, it is really just marketing driven. Facebook/OMF has already, essentially overnight, occupied the âopen map dataâ mindshare, it doesnât really matter if the data is usable or not.
11 Likes
It could be helpful not boasting that our data is difficult to use.
2 Likes
Iâve been around on and off since 2009. Obviously I cannot speak for Overture, just share my thoughts.
-
Customer orientation
OSMs stance has largely been âwe collect the data, let the data consumers sort it outâ. As a consequence, OSM doesnât have customers, just people using our data. No brand recognition or loyalty either.Critically, the emphasis has shifted - from building the map to maintaining and providing it.
We could look into a) what data our users actually value and need more of / better quality, and b) offer a (paid?) option where data excerpts with defined meaning and structure are provided and maintained with the junk weeded out. Oh, waitâŠ
-
Innovation
Essentially, we are still in 2009-2015. But those goals have been (over)achieved, and the world has moved on. Remote sensing, satellite navigation, OGD, live data feeds and AI are commonplace. I donât see any of this reflected in OSM. Advances are happening with tags, tools and applications, but not at the core.Nowadays, OSMs sheer size and complexity make any substantial change a huge undertaking beyond the capabilities and funds of any of the participants (not to mention the decade of debate that would ensue).
-
Reliability
As it stands, OSM is a bunch of volunteers who could turn to mapping Mars if the opportunity arose. The community is still struggling with an image of amateurs and socially challenged nerds. Thatâs not the kind of partner a professional data consumer would want. -
Unresolved internal conflicts and contradictions
Going way back to the not to be questioned ideals of the founding fathers (were there any mothers?), just to name a few:- Collect data vs provide something useful
- Community orientation vs product orientation
- Free data, but with a not-so-free license
- Monetization, but only for some
- A steering body where nobody should have a decisive influence
- Brand recognition, but only through intermediaries
- No rules, unless the Wiki (subject to change without notice), the Community (the handful of guys who happen to be online), the DWG or your favourite deity say otherwise
- Then again, its a do-ocracy, make your own rules
- Then again, none of the above may be applicable
- No mass edits or imports, except manually, or under the radar, or if nobody notices or cares, or if you ask nicely
- Use QA tools, but donât rely on them; after all, there are no rules and everything is subject to debate
- Play in major league, but depend on donations, but not in a way that someone might want to influence us for their own benefit
-
Efficiency
Human mapping on the ground has its advantages (apart from being fun). That said, there simply arenât enough experienced mappers around to even maintain the ever-growing amount of data, let alone extend it. Itâs a battle we canât win. The structures and tools for data management simply arenât there.Our strength used to be in numbers, but how long that will hold is anybodyâs guess.
Overtures alpha release may be junk, but with their infrastructure and reliability percentage they have the means in place for continuous, manual and automated refinement. Just a matter of time, money and patience. The big question ist, how much will it take to catch up with millions of unpaid mapping hours world-wide? (nothing, if they can re-publish our data)
-
Leadership
OSM of 2009 is over in most respects. Whatâs needed is a new vision from now onwards. Who will participate and why? What will they strive for, and for whom? Which rules and tools will they need? How will they be motivated? Where will they get resources? What makes OSM different and better? How will it be recognized (Intel InsideÂź, anyone?) And whatâs the policy with respect to other players - compete, co-exist or cooperate?The OSMF is doing what it can, but it is no match for professional, commercial, focused leadership - unless Overture trips over its own feet, or runs out of steam.
17 Likes
Thinking about it, a strategic partnership with a big data consumer might help to set priorities and get brand recognition (and resources) as well.
1 Like
Yes, I think you touched on right points. OSM as of now is an âanarchyâ with nobody to refer to in case of stalemates. Too little cooperation between different local communities, core OSM software devs (openstreetmap-website, openstreetmap-carto, iD, Nominatim) and data consumers.
That said, I would be far from writing off OSM. If anything, I hope itâs gona be an impulse to re-invent ourselves. And nobody said we couldnât do what Overture members do, but more efficiently. For example, while AI has been used to trace buildings with varied results, NorthCrab just pioneered using AI to cross-check Polish government dataset (BDOT10k) with orthoimagery before importing buildings. Now he moved on to adding pedestrian crossings.
What we also have is an army of people with local knowledge, willing to check stuff on the ground. I highly doubt Overture will deploy yet another fleet of âstreetviewâ vehicles.
7 Likes
Was remarked earlier in this thread along the lines of us volunteer OSM mappers made mules put before the OMF cart.
Simon, it might be the first time we seem to disagree! Quality has EVERYTHING to do with success. In the short term, yes, it is possible for âmammoth marketingâ to âoverwhelmâ the truth of truly shitty quality.
But in the long run, shitty quality is flushed into the sewer and the quality stuff keeps getting reached for more and more often.
Map data are ALL ABOUT âusability.â Sure, money and techy flimflam dissolved around the lies and commercial nature of âthose sortsâ of social media (btw, Iâll call this Discourse right here OSMâs version of social media, although there are Slack communities and moreâŠ) can dupe people for a while. The USA had an ex-president (not 45!) once say âyou can fool some of the people all of the time, but you cannot fool all of the people all of the time.â
Shitty quality (yeah, Iâm calling it what it is, when it is shitty) is foolish. People might put up with it for a while, or find themselves (temporarily, one hopes) locked into or forced into it for a while, but theyâll reach for the quality good stuff as soon as possible. OSM intends to be there.
3 Likes