Hacker News Reader

Comment by crazygringo | Hacker News Reader

This article makes zero sense to me.

It's arguing that corporations are contributing data to OSM, but that this somehow... changes the proportion of what kinds of things OSM covers? That traditional users have added "hyperlocal features, such as a neighborhood bench or informal walking path" while coporations are adding "roads, buildings, and all other points of interest" and "a particular emphasis on improving road data".

How is this anything but good? As long as data is correct, the more data the better -- proportions are meaningless. Buildings and roads aren't taking away from benches and paths. And I'm pretty sure buildings and roads are benefiting all users anyways, no matter who contributes them.

I'm utterly baffled by what logic the author imagines this is a bad thing. The data is open, so it's not like corporations are winding up with exclusive ownership of anything.

So what's the problem...?

View on HN

One of the typical ways this manifests itself in reality is a dedicated local mapper adding or correcting a detail where they know that some common data sources are either wrong or out of date. At some point later a corporate mapper the other side of the planet working from imagery will come along and change it back to agree with the imagery they have, despite it being 4 years old. Local mappers get exhausted from the effort it takes to keep their patch from being broken and disenfranchised with their edits being drowned out by the noise of the corporate machine.

A devils advocate answer of course, I'm personally on the fence on the issue.

To also play devil's advocate, that could happen with a non-corporate mapper on the other side of the world, too.

The solution, I think, would be a mandatory date field specifying when the data was accurate, and only allowing updates with more recent data.

It could happen, but it generally doesn't.

That's the kind of thing that happens when someone is paid pennies per edit and given insufficient resources to guide those edits, which really only happens when someone points a hose of money at a contractor.

I've long been a fan of a mandatory time field! Now to get Stack Overflow to implement it too :)

That, my friends, would be the first step for my dream to finally map historical data, like in the OHM, directly in the OSM without anyone feeling bothered by it.

It's happened to me. Not just on maps, but on Wikipedia.

That hasn’t been seen in practice though. Apple has been making changes from surveying with their own vehicles, not out of date imagery. Amazon makes updates from delivery drivers reporting issues making deliveries.

This specific thing has been seen in practice, it was fairly regular occurrence back when I contributed to OSM and a source of lots of similar complaints as this article at the time. In fact, often the argument was framed as automated vs human editors as some non-corporate large-scale data imports had gone badly wrong too.

Even at that time though, it seemed like the corporate support was, on balance, a positive thing. The main worry at that time (for me anyway) was some corp getting smart and pitching themselves as a middleman for other big corps. Luckily the contenders all got slurped up by big corps in a feast of mutually assured destruction, and since no one could trust their future to be built on a competitor's playground, they seem to have all defaulted to OSM.

Similar to Linux, having a diverse set of medium sized corps on your team (and maybe a few really big ones that see you as a complement, not a competitor) seems to be a winning strategy.

I think the concern is big companies might get into governing positions and make decisions about the projects future... For example, they could decide to divert some efforts into a commercially licensed 'extended' dataset, or they could sign restrictive deals to get hold of 3d laser scan or satellite/aerial imagery that isn't so easy for volunteers to collect.

Thanks. I guess I just ask, is there any basis for that fear though?

The OpenStreetMap Foundation's governance [1] isn't tied to who contributes data. Is there any evidence they are being "corrupted" in any way at all by corporations adding data?

And like any open project, if its governance were captured by "bad guys", users can always fork it and "vote with their feet" since the data's open.

If OSM really was being subject to a hostile board takeover by corporations then obviously that would be hugely newsworthy and concerning. But it's like the article is trying to stoke fears of this while presenting zero actual facts to support it.

[1] https://en.wikipedia.org/wiki/OpenStreetMap_Foundation#Gover...

Based on previous large company involvement in 'open' projects, yes. It usually starts out fluffy and friendly and just 'wanting to help'. The large company (or companies) take on an increasingly large role eventually leading to taking on and then over governance... then it's often game over as far as the original intent of the project. By the time they are in the governance roles, the dissenters have typically already been marginalized and/or are gone.

This is a very old strategy. Companies (especially the large, name-brand ones) aren't getting involved because they believe in the mission/objectives of OSM but rather that they see a strategic advantage in their involvement. They dip their toes in, get entrenched and then make their move. It's a lot faster and less expensive for them than trying to replicate what the project has produced.

I'm very interested. Can you point to specific examples?

And did the original contributors fork the project to a new successful initiative? Or if not, why not?

I'm also curious why companies feel the need for this -- if the data is all open in the first place, why do the companies even need to bother? Why don't they just use the data directly?

Without specific examples it's hard to understand.

Specifically related to the story, Apple has a rather troubled history as an Open Source 'contributor'. They tend to not contribute unless they need to and then prefer to run the show. They basically steamrolled KHTML with a fork of their own code a.k.a. WebKit (which they effectively control development of exclusively). Imagine them doing something like that to OSM... which they will if it ever suits their purposes. See most other Open Source projects Apple has been involved in... it's not pretty. So yeah, Apple playing a significant role in OSM should have alarm bells ringing in the heads of other contributors.

That's not to say corporate involvement/sponsorship isn't important: it's vital to many important projects. The issue is when companies start trying to influence the projects beyond their technical needs and start trying to wrap their business model around them (a good indicator of that seems to be their involvement in governance) that the trouble starts. I think that's the concern being expressed about OSM.

That would be a legitimate concern but the article focuses so much on stuff like "look how much edits Apple made!!" that it's really hard to tell if it was a concern for the author, at all; or if there are others (e.g. it spends about as much time talking about the governance concern as it talks about diversity in the governance body; so it's unclear if author is worried that companies might take over, or is just unpleased with the governance in general).

Same as the parent poster, I'm baffled. This article makes no sense to me. If it had a point to made, it buried it in thousands of words of blabbering.

If we look very well at it we see Apple made the most contributions to map Africa. Who's going to complain about that? Google, maybe? I guess african roads don't suit the google maps cars.

It's also a curious business decision... Apple doesn't make much revenue in Africa, nor have many users there. Why direct all their mapping efforts there?

No-one said that Apple directs all (or even most, or even just a significant proportion) of their mapping efforts there. Should they ignore a continent?

It makes sense to commoditize their complement.

Diversity and inclusion?

Africa is the largest growing continent and just like China has been an extremely important market for many companies, so too will many developing African countries. Apple would love to be the market leader there and any headstart they can get will pay dividends in the future.

But Apple products are generally extremely expensive ( usually at least multiples of the cheapest equivalents), and that kind of thing doesn't sell well in underdeveloped developing nations, which cover most of Africa. Just look at median income and GDP PPP in China, and any African country. The mere fact that China is actively (kindof) neocolonising multiple different African countries should be telling as well.

Yeah and I guess American roads don't suit the google maps cars either.

I mean just look at them: https://i1.wp.com/thirdeyemom.com/wp-content/uploads/2017/08...

That seems to be in Haiti: https://thirdeyemom.com/2017/08/02/what-it-is-like-to-drive-... (it's in the first slideshow).

Yes exactly, a American country with terrible roads.

How can Google map roads in the Americas with their terrible roads?

We don't make distinctions between any African countries anymore so I think it's only fair we start treating the other continents the same way.

(I mean of parts the US are close enough to third world countries anyways, 11 year olds dying in mobile homes because corrupt leaders hobbled basic utilities and all)

"American" implies it's in the USA, like if someone is described as "American", they're from the USA. European, Asian, African, Australian, North American, and South American all refer to the continent, Antarctican seems to be a similar word (without much use, unsurprisingly), but I don't think there is one for the whole Americas in English, because "American" isn't it.

Someone is from the Americas I'm going to call them American. What say do people from the continent get anyways?

They're all the same sh*thole countries anyways with corrupt leaders full of murders and rapists right?

One let a TV clown run them into the ground, another is run by drug cartels, awful continent.

If I, a Ghanaian, have to constantly go by "African", then the US being lumped in with Hati shouldn't be inaccurate at all.

Weird you thought otherwise.

I'm not American, or from the Americas, I'm just letting you know the meaning native English speakers will take from the word. Call non-Americans American if you want, but don't be surprised when you're misunderstood.

You should learn the meaning of the word "facetious".

And I'm not sure what makes you think I'm not a native speaker of English.

Apparently you don't even understand blatant sarcasm, are you one?

> You should learn the meaning of the word "facetious".

I know it well.

> And I'm not sure what makes you think I'm not a native speaker of English.

You apparently don't know the definition of "American". I have never seen a native speaker make a mistake like that.

> Apparently you don't even understand blatant sarcasm, are you one?

I understand sarcasm just fine. Calling a Haitian road "American" is not sarcasm. Yes, English is my first and only language.

You said a picture was from the USA, and I doubted this, so I went and spent time chasing down the actual source. That's not facetiousness or sarcasm, that's either an innocent error, or misleading and antisocial. I interpreted it as innocent error, and attempted to politely correct your knowledge and explain, and now you reply with this argumentative comment, implying you deliberately attempted to mislead and waste our time. I won't reply further.

You're just too dense to understand what was said, you shouldn't have replied in the first place.

Person implies somehow Africa, the continent, is somehow unsuitable for Google maps vehicles (despite the fact Google has mapped hundreds of tens of thousands of miles in multiple countries in the continent)

This comes from treating all of Africa like it's one giant country with the same infrastructure everywhere.

I facetiously (a word you do NOT understand apparently) show a picture of Haiti, a North American country and say, since their roads are terrible all North American roads are terrible, and refer to an entire continent as if it were a single country that includes the US.

From here your subpar English skills fail you, to the point I'm forced to use a word somewhat similar to facetious but not quite as accurate (sarcasm) because the proper word is above your understanding.

Sometimes being imprecise allows a person with a challenged vocabulary such as yourself to catch up, but of course the Dunning-Kruger effect is in full force here, so you use it as an assertation that you're not the one failing to understand basic vocabulary.

It's a shame English is your only language and yet mastery of it escapes you.

There's no other word for people from the United States because of the weird naming of the country.

Canada - Canadians, Mexico - Mexicans, Brazil - Brazilians.

United States - ......United Statesians?

That's like people from the DRC calling themselves 'Democratic Republic of the Congolesians.' Nope, they just shortened it to Congolese. The United States of America just did the same thing. 'Americans.'

They all use it, and for the most part so do their neighbors.

Until some clever soul figures out a better way for the bologna of the North American sandwich to refer to it's citizens, the largest collection of English speakers on the planet will probably keep on trucking with the same boring term.

This unitedstatesian exists in other languages. In Spanish: estadounidense (in broad use). In French: étatsunien (rarer).

> the largest collection of English speakers on the planet

Indians?

I wikipedia'd it before I posted.

https://en.wikipedia.org/wiki/List_of_countries_by_English-s...

United States - Total English Speakers: 283,160,411

India - Total English Speakers: 246,344,737

In that article the Indian pop is from the 2001 census at 1,028 M. Current population is 1,388 M, keeping the same percentage of English speakers would place it at +305M speakers, although doing the same for USA also place it at +310M.

That being said I think the table is wrong because the numbers don't add up to the total and also don't match the source. Based on that census India should have only 12.18% of English speakers, not 22.18%, so you should be correct in any cases.

It's incredible how dense people can be when their world view is so narrow.

The point isn't there should be a word for it, the point is Africa is not a damn country. Such obvious facetiousness but thick skulls can only fixate on "wHuT amErIcAnS Are thE US"

South Africa has as much in common Sudan as the US has with Haiti

There are plenty of countries on the continent than can (and have been) mapped by Google Maps.

Saying that Africa is somehow inhospitable to Google Maps is exactly like acting like North America is inhospitable to Google Maps because Haiti has bad roads.

Hell I'd take driving on the roads in that Haiti picture over some in places like "Chiraq", probably less chance of a bad outcome in a Google Maps car.

US(-Amer)icans.

Is there really precedent for companies taking over open-source upstreams and turning them proprietary? I expect the corporate users appreciate the price & freedom at least as much as everyone else.

MySQL springs to mind

Oracle took over MySQL as the acquirer of Sun, not as a heavy MySQL user with too much influence.

Also mysql _sold_ their governance to Sun - it wasn't _taken over_.

Not sure what "open-source upstream" means, but in software, there is the complaint that companies like Google can add a lot of complexity to software projects, which serves Google, but not the classic home user of Linux. That can make it effectively their own corporate project and drown out more casual developers.

That much is true - direction and feature set. The accusation that seems weird is that companies would be motivated to make it proprietary just because they use it. At most, it seems like you’d get a niche for “enterprise” distributions with support contracts on the same stuff.

Ironically OSM is the opposite. The hobbyist mappers do bus routes and admin boundaries and mountain trails and all manner of complex stuff. The corporate mappers are really just interested in the most basic element: drivable roads.

Oracle is the most (in)famous via various acquisitions. But IBM/RedHat/Canonical/Microsoft are a few other large players that have been trying with varying degrees of success recently related to Linux. Corporations appreciate ever increasing EPS and decreasing costs... nothing more.

For me this discussion is new, so I don't have good answers to these questions.

One thing that happened when Pokemon Go got popular is that people (users) were conducting vandalism, in the sense that many pokemon figures were added to the main map of OSM. People have to work really hard to remove these, and they might burn out on this. There is already a lot of vandalism, like people adding their friends drinking shed, or whatever they think of as funny.

Another thing that already happens with volunteers, is people thinking their working area is "their area", while no such concept exists. This can already give a lot of friction among volunteers. I can imagine if billions in money are at stake, big companies with a massive crowd of emplyees might want to really grab those working areas and drown out volunteers. If that gets a big problem, volunteers that were active for years might drop out.

Oh, one other thing. Many volunteers are maintaining and guarding hiking routes and cycling routes. These get damaged easily (I just did that yesterday myself) and these volunteers sometimes carry a big burden. The big companies don't have an interest in these routes "by the people, for the people", there is no money in that. If they get more often damaged, it will not be regarded as funny by these important long-lasting volunteers.

Is this virtual or real damage? As in, the trail route is imperfect and misleading about the actual trail, or the trail is revealed via the map and others come and physical use damages the trail. The second reminds me of land owners sneakily putting "no trespassing" on land that is legally publicly accessible (I think this happened in Sausalito about a year ago)

There might be many situations, but I was mostly talking about virtual vandalism. When you edit points, lines or routes in the wrong way, they get damaged. The volunteers guarding these need to fix them themselves, which can be a burden and they can get emotional already. You can imagine how that goes when more damage gets done where the fixing is left for those volunteers.

I don't know much about land-owners doing vandalism, often people try to communicate, sometimes this works out okay.

It's often as simple as mis-tagging. I do a lot of mapping of mountain bike trails (think narrow, hiking-type trails which are also open to bikes) and these are designated as paths (highway=path), open to cycles (bicycle=yes or bicycle=designated) and some mountain biking-specific tags.

What'll often happen is a well-meaning person will come by, often using an editor that doesn't differentiate between types of bicycling, and change the route to be tagged highway=cycleway. This is the tag for things like paved bike paths, such as urban cycling infrastructure.

The result is now these mountain bike trails, narrow/rocky/tight/steep stuff, shows up in tools that offer bike commuters directions. Just... not great.

So many of us keep an eye on the stuff in our area to see that the tags stay as they should.

Is this really such a problem? I guess this is my first time hearing of it, but I believe OSM allows you to mark areas where public wifi (or private wifi, in the case of businesses that provide wifi to customers), is available. Wifi access is not a tangible, physical thing (even though of course the physical layer is), but OSM provides a framework to map it because it's tied to a location, and the information might be useful to some people.

I am not sure what you are replying to, so I assume there is some miscommunication.

I don't know anything about mapping public Wifi access, like in shopping centres.

What I meant is that already some people feel like their city and a 30km radius around it is "their area" and they want to control what and how things get mapped there. It can lead to heated and emotional discussions, we are people after all :) If a big company with lots of human resources would do this, there is hardly any fighting back.

I was replying to the Pokemon capture points added to OSM being considered vandalism.

I think the issue was that Pokemon Go used certain types of map elements as their Pokemon capture points, e.g., Landmarks, or something. People weren't adding "Pokemon capture points", they were adding non-existent Landmarks so that they could be imported to the game for their own advantage.

Ahh, thanks for explaining. That's definitely an abuse of the service

OSM contributor here: agreed with you, not sure what they're on about. All the sources are also old and known, with the newest references to the latest State of the Map conference. It's a long article with a lot of random historic fun facts, but I don't quite get the point.

Thank you for the summary. I couldn't get past the popovers and the broken scroll to read the article.

Reply on news.ycombinator.com