Overture Maps

[rschen7754]

Not really sure what to make of this but it seems that Linux, Amazon, Meta, and Microsoft are teaming up against Google Maps.

https://arstechnica.com/gadgets/2022/12/linux-amazon-meta-and-microsoft-want-to-break-the-google-maps-monopoly/

[Scott5114]

Sounds like Microsoft/Amazon/Meta/TomTom wanted to do something with OSM but without having to deal with that pesky OSM community.

How many millions does it take to buy some damn shields for the base map?

[jeffandnicole]

Google Maps is hardly a monopoly; they just have such a strong product that most people think of it first when they need a travel aid.

If anything, there's still lots of ways to improve that product by offering different directions based on the vehicle in use. Say, a truck option that would route certain vehicles away from low clearance or weight excluded bridges.

The open source world does not know roads very well, and everyone thinks they are the best driver. Giving every John and Jane Doe the opportunity to "fix" a map, even though they can't tell an interstate from a dirt road, will only lead to disaster.  Get certain roads involved, let's say bicycle coalitions, will have them continously "closing" roads, leading to tremendous confusion.

Can't wait to see what the competition provides!

[epzik8]

Yep, Google and Bing can only handle so much.

[vdeane]

Sounds like Microsoft/Amazon/Meta/TomTom wanted to do something with OSM but without having to deal with that pesky OSM community.

How many millions does it take to buy some damn shields for the base map?

I think a comment on the Ars article put it well:

My experience with (trying to) contribute to OSM is that there's a lot of stuff OSM explicitly doesn't want. As this press release says (press release in accuracy shocker!), OSM is really kind of a community of map nerds. For instance, I tried to contribute a bunch of street numbers for my city which I'd taken from the city's own sources - after getting the city to publish them under an acceptable license - and doing manual sanity checks and so on. After a bit of debate, OSM rejected them, because apparently they only want street numbers you collect by hand by walking around the city with a GSM tracker. Because that's...what they want, I guess? Importing open data from somewhere else is not okay. Or wasn't at the time.

I believe they also explicitly don't want things like information on commercial businesses. You know, things like names and phone numbers and menus from restaurants.

Which, I mean, cool, OK. Their project, their rules. But if they explicitly don't want useful data, then naturally other projects are going to build on top of the data they do want. Obviously this effort can't just use and contribute to OSM - it's going to want street numbers. And business names. And, you know, things of that nature. Which is stuff OSM apparently just doesn't want.

Sure, us map nerds might like just panning around looking at the roads, and only really have a complaint about the lack of shields, but lots of people use other features that are either half-baked or not available at all in OSM.  Driving directions is a big one.  I just did a test run between my apartment and my parents.  I couldn't even type in my apartment's address.  Had to use the dummy address nearby (which I usually use with Google, since it did odd things with my actual address when I first moved here and the habit stuck) or drop the point manually.  Not having the list that appears when typing is also an annoyance.  It also doesn't let one drag the route (something which I use all the time, as I use directions for estimating travel time and saving routes for later, not for actually figuring out how to get somewhere), and the drive time listed is an hour longer than the trip actually takes (Google is usually right on the dot with how I drive, though an option that wouldn't go faster than the speed limit on any route sections would be nice - not only because I expect that I'll one day have to drive the limit exactly with the proliferation of speed cameras, speed limiters, and the like, but because there are some areas like around DC where traffic flows a lot faster than my usual "speed limit + 5-7 mph" travel speed).

Not to mention the things the comment mentions on businesses.  I use that data.  If I want to get a website for a business, or their hours, I turn to Google Maps.  I also love using Google Maps to get photos of businesses and their products, as well as reviews.  I would miss that if I moved my main map browsing somewhere else.

Street view is another big one.  Every other product that has attempted it is either less pervasive, doesn't update often (or at all), isn't panoramic, and/or has other interface issues.  It's not just fun to look at - it's how I make sure I know where I'm going when driving, by checking out every single turn I'll be making in advance (where possible).

[Scott5114]

I don't really consider a lack of business information a deal-killer for a mapping service. That's not on state DOT maps, so I've never felt like I need to look it up on Google Maps. It is a handy shortcut that I'll use if I'm, say, staying in an unfamiliar hotel and I'm wanting to pan around and see what food options are nearby. But for stuff near home I'll just put the business's name in Google search if I want their website or hours. Reviews aren't really helpful to me because I'm pretty good at coming up with my own unfounded opinions.

Lack of driving directions is a big missing feature in OSM, though. Heck, search leaves a lot to be desired. And it is disappointing to not have a Street View equivalent, but there's really no feasible way for an open-source product to acquire that data. (I kind of doubt it's feasible for Google to acquire that data, to be honest. The only way it makes sense is as a loss-leader to get people to look at Google Maps ads, and any day now I'm half-expecting them to hear that they've done away with it to cut costs.)

[GaryV]

I've put business names into Google Maps as a shortcut to looking them up on their website. For example, if I have reservations at Fancy Inn in Anytown, I'd have to look up the inn's website to find the exact address. I can look it up by name on Google Maps and find how to get there.

[Scott5114]

Generally, I find that I'm better served looking it up on the business website anyway–businesses frequently update their hours on the website but not Google Maps, there is additional information like menus or terms and conditions that don't fit the Google Maps template and so are not displayed, etc.

[skluth]

I saw this a few days ago on Gizmodo. That article says much the same thing. All I have to say is good luck to them. But I have absolutely no faith in their ability to pull it off, mostly because of Microsoft and FB/Meta's being involved.

[hbelkins]

I'm an iPhone user since the first one came out, as I've been a Mac user since 1987, and my default navigation source on the phone is Google, not Apple. If I'm not mistaken, you can plot a route on the desktop version of Google Maps and port it to your device using your Google account. I don't know how to do that with Apple Maps since there is no Apple Maps desktop app or Web site.

Also, I think that directions generated from the desktop version of Google Maps could be sent to your OnStar receiver in your GM vehicle.

[Rothman]

At least for Google for Android, complicated desktop routes cannot be imported (e.g., with via points and the like).  It is highly frustrating.

[Scott5114]

I import complicated routes into my car by writing them on a post-it.

[LilianaUwU]

I can tell you what Microsoft, Facebook, Amazon and Linux are trying to do here will flop. There's no way they will break the Google Maps monopoly.

[Scott5114]

I can tell you what Microsoft, Facebook, Amazon and Linux are trying to do here will flop. There's no way they will break the Google Maps monopoly.

I don't know if I'd be that confident in their failure. If the Linux Foundation is involved, that means the end result will be open-source. If you can get it to even 80% of the feature set of Google Maps, not having to pay Google for licensing nor deal with any of their API horseshit will be enough to tip the scales in its favor for enough use cases to make the project worth it.

Speaking in the general sense, whenever I'm choosing software, I habitually sacrifice features for the ability to control my own data and/or keep costs (system resources and financial) down. Not everyone is willing to make that sacrifice but it's nice for there to be an option to do so if it makes sense for your use case.

[Rothman]

I can tell you what Microsoft, Facebook, Amazon and Linux are trying to do here will flop. There's no way they will break the Google Maps monopoly.

I don't know if I'd be that confident in their failure. If the Linux Foundation is involved, that means the end result will be open-source. If you can get it to even 80% of the feature set of Google Maps, not having to pay Google for licensing nor deal with any of their API horseshit will be enough to tip the scales in its favor for enough use cases to make the project worth it.

Speaking in the general sense, whenever I'm choosing software, I habitually sacrifice features for the ability to control my own data and/or keep costs (system resources and financial) down. Not everyone is willing to make that sacrifice but it's nice for there to be an option to do so if it makes sense for your use case.

I wonder about the quality of their traffic info compared to Google/Waze.  That's a huge hurdle to overcome.

[Scott5114]

I'm imagining this is going to go like pretty much every open source project–implement the 80% of features that are easy and leave out the 20% that are hard or expensive. Fortunately, most of the basic features capture most use cases, so it will be a suitable replacement for most people if they're open to it, but if you're one of those people that demands 120% of the feature set of the corporate software plus requiring it to make coffee and wipe your ass for you, you'll probably have a bad time.

Personally, if a map existed that had attractive cartography with county lines and good wayfinding and search that'd be enough to take care of the majority of what I use Google Maps for. Not having Street View would be sad but I don't expect an open source project to have the resources to collect that. And I don't use the traffic or business data at all so I don't care if that's there or not.   

[Rothman]

Interesting.  I find Google Maps' use of traffic data in routing to be invaluable -- even when I decide to not follow it.

[Scott5114]

Oklahoma City's trouble spots are predictable enough in times and locations that I have no need to look up routes around them. If there is a surprise backup due to an accident or whatever, the detour will invariably be "exit to the surface street grid, jump over a mile, and run parallel to the freeway until you think you're past the backup". (This strategy seems like it works pretty well in general for cities based on a grid–I tried it in Las Vegas when the cops pulled the boneheaded move of closing down I-215 for a police officer's funeral, and was able to muddle through despite not having much familiarity with the city.)

[bulldog1979]

I'm an iPhone user since the first one came out, as I've been a Mac user since 1987, and my default navigation source on the phone is Google, not Apple. If I'm not mistaken, you can plot a route on the desktop version of Google Maps and port it to your device using your Google account. I don't know how to do that with Apple Maps since there is no Apple Maps desktop app or Web site.

Also, I think that directions generated from the desktop version of Google Maps could be sent to your OnStar receiver in your GM vehicle.

Current versions of MacOS have a Maps program on the desktop, and information from it can be sent to an iPhone. I don't use it that often, and if I need directions on my phone, I just initial the search on my phone to start.

[Rothman]

Oklahoma City's trouble spots are predictable enough in times and locations that I have no need to look up routes around them. If there is a surprise backup due to an accident or whatever, the detour will invariably be "exit to the surface street grid, jump over a mile, and run parallel to the freeway until you think you're past the backup". (This strategy seems like it works pretty well in general for cities based on a grid–I tried it in Las Vegas when the cops pulled the boneheaded move of closing down I-215 for a police officer's funeral, and was able to muddle through despite not having much familiarity with the city.)

Well, I suppose that works for urban areas to a degree (although I snicker what that approach would do in Syracuse -- greatly take more time than just staying in the backup in a lot of cases).  Do you take long road trips often?

I'm thinking of the last total solar eclipse, where traffic in KY and TN went all nuts after it was over.  Finding a way around the masses was made a whole lot easier by using Google Maps, even by just knowing where it was sending people to come up with my own detour (got into my destination hours before my parents who followed Google strictly).

[Scott5114]

Generally, I use paper maps when I'm on the road if I can help it. Much easier than trying to refer to the phone if I'm behind the wheel (you usually don't have to worry about the map closing itself, and when it does you don't have to fuss around with unlocking it, and the need to "pan" is much less frequent because most state-DOT-style maps include more detail than the equivalent zoom level on digital maps).

Granted, that does nothing to help with backups in rural areas, but you have to admit that situations like the eclipse you mentioned aren't terribly common. Much of the time, getting off the freeway in a rural area to circumvent a backup isn't going to help much anyway and just result in a lengthy detour, especially if you don't trust the county roads.

[Rothman]

Using paper maps is much more distracting from driving than having my phone running a map app.

Sure, the eclipse was just one extreme example.  Still, it could be applied to when there are backups outside of urban areas.  Also reminds me of the time I-95 was completely jammed somewhere in southern NH with incredible delay times (another data report that is useful that you're not getting from a paper map).  Gave me the info I needed to choose a different way given the time trade-off.

[Scott5114]

I don't really see how a paper map could be more distracting. Before I put the car into gear, I unfold it, put it on the seat, and glance over when I need to.

With the phone, I glance over and see a blank black screen. So now I have to press the unlock button, draw the unlock pattern on it, hope that the last thing I used on it was the map and if not hunt around on the home screen to find the map program icon, pan the map to where I am, zoom in or out to see what what I need to see...

[skluth]

The traffic info shouldn't be too difficult to collect; it just requires permissions by users to send traffic data to a central source. The problems are who holds the data, who can use the data, how the data is used, who updates the data and how is it updated, etc. Google spends a fair amount on all aspects of Google Maps; it's not just GSV. Their users (both Maps and Waze) constantly contribute traffic data in real-time. They have a centralized system for updating from contributors. They work with Digital Globe and others to keep their imagery current even in many areas with few people. I'm not worried about the Linux Foundation wanting to do the right thing. I just don't trust Meta or Microsoft to play well with the others as their past history is to stovepipe what they want into their own environments.

[hotdogPi]

I just don't trust Meta or Microsoft to play well with the others as their past history is to stovepipe what they want into their own environments.

Microsoft bought out Minecraft and didn't run it to the ground. Meta, on the other hand...