Hacker News new | past | comments | ask | show | jobs | submit login
Apple’s New Map, Expansion #5: Northeast U.S. (justinobeirne.com)
81 points by l1n 13 days ago | hide | past | web | favorite | 41 comments





Wow, this guy makes super in-depth posts. This one about Google's methods for creating "Areas of Interest" is fascinating: https://www.justinobeirne.com/google-maps-moat

I wonder what kind of change resulted in that first before/after image, where Allentown PA disappeared at that scale and Philadelphia appeared. It seems more right than before, but also, why not both?

Also wonder why some town names are in the same exact location, but others move just a few pixels. Updates in annotations, or in a constraint solver?

Cartography is such an interesting topic. I experimented using Douglas-Peucker on 3D images of protein molecules in grad school–cool stuff. I find myself often using cartographic principles in analogy to software architecture and codebase organization.


Three cities for CT, only one for NJ. Syracuse doesn't merit a label despite plenty of room to relocate NY.

I’m not sure what they mean by “top 9 urban areas” the top 9 metro statistical areas in the US for 2018 [0] are: NY, LA, Chicago, Dallas, Houston, DC, Miami, Philadelphia, Atlanta.

Apple has 6 out of 9, not 8 out of 9. San Fran and Boston are not top 9. Boston is 10, SF is 12.

They still have a lot of coverage but they can at least say top 10 or something.

[0] https://en.wikipedia.org/wiki/List_of_metropolitan_statistic...


The "Largest U.S. Urban Regions" chart is likely referring to the U.S.'s 569 "Primary Statistical Areas": https://en.wikipedia.org/wiki/Statistical_area_(United_State...

Interestingly, while the 3D buildings are present in the vector map of Washington DC, they appear to be disabled in the Satellite view. We wondered if this was intentional obfuscation, in the same vein as how the rooftops of certain DC buildings were blurred out in early satellite imagery on Google Maps.

EDIT: clarification


Interestingly, while the 3D buildings are present in the vector map of Washington DC, they appear to be disabled in the Satellite view. We wondered if this was intentional obfuscation

Probably not. It seems to vary based on the availability of data.

In fact, it can even vary from hour to hour. Over the weekend, the 3D satellite of downtown Chicago disappeared for a while, while the 3D vector was still available. It restored itself in a few hours.


Good in-depth article. Only nitpick is the part about "Apple’s rollout might be slowing." They have to really (in my opinion) misrepresent the data to say that it's slowing. The author quotes how the latest expansion is smaller than the last expansion. However, they don't take into account that the "Expansion 5" came only one month after "Expansion 4", which came /five/ months after "Expansion 3". If anything, it would feel like the rollout is speeding up, since it came hot on the heels of the previous expansion.

The author does, at least, qualify that the most recent expansion covers a lot more population density than previous expansions.


It's a bit harsh to say that the data is "misrepresented" here, especially since it's immediately qualified with the population density info.

Apple could very well be slowing down, and it's not immediately clear to me that they'll actually make their deadline.


I don't think it's harsh. The author says that the rollout might be slowing. They justify this with smaller increase in coverage in the most recent update. But the updates are not spaced equal amounts of time apart. So the actual expansion per unit time is continuing to increase.

Compare to OS releases. Windows 10 gets updates every month, with feature updates every six months. Compared to before, these updates have fewer features. But they are coming much more frequently. Can we say updates are slowing, because they don't have as many features in each update? Or do we have to look at the features per unit time?

I agree that it's not necessarily clear that they'll make their end of 2019 deadline.


I disagree. The first few expansions were increasing at a rate of 5x and were roughly 2 months apart. Then the 4th expansion took almost six months and was only 1.4x bigger. And now this latest expansion, which took ~2 months, is smaller than the last two. So it's definitely fair to speculate that Apple might be slowing.

I'm curious what Apple has to actually gain by doing Maps?

Is it just about not wanting to share user information with Google as a selling point for their devices? Competition for Google Maps is absolutely a good thing to spur innovation.

At this point, I'm still using Google Maps because they still seem far more usable and feature filled. I'd love to see a viable alternative.


Maps is a required feature of smartphones. If Apple does not have Apple Maps Google has leverage over Apple.

The reason Apple started Apple Maps was that Google refused to allow turn by turn navigation and vector maps on iPhone and required location data from iPhone users. Only when Apple Maps was introduced did Google release these features and they lost access to most location data.


Everytime I see an article on HN about Apple Maps, I get hopeful that there's an announcement buried that they're releasing an Android version...

I agree that Apple (& Bing/MS) is vital... But in order for the goog to be toppled, there needs to be apps on all platforms. It seems ludicrous they want to redistrict their map and messaging to iOS only, but there's a real opportunity for Apple to spread out into competing platforms.


There is Apple MapKit to embed into websites https://developer.apple.com/maps/web/

I'm already using that for a couple of projects. It's not the same though. (PS: It's got a great trucking api)

> I'm curious what Apple has to actually gain by doing Maps?

The same reason Google started making hardware for Android. A company should try to control its core competencies[1]. The intersection of software and hardware is a core competency for Apple. At the application level, navigation is a critical piece of that software.

[1] https://en.wikipedia.org/wiki/Core_competency


What is the value (to Apple) in providing this kind of detail in maps, other than aesthetics? I'm trying to understand why Apple (or any company) would invest money into what seems like a marginal improvement in UX. The city landscape is constantly changing, so there's associated maintainence cost too.

This was previously covered by a well-known post by the OP:

https://www.justinobeirne.com/google-maps-moat

The necessary link between aerial imagery and on-ground navigation by mobile phone is a 3d map.

I'd like to know where the front entrance is vs. the side entrance, is there a ramp vs. stairs, are these two buildings connected or not, is there a skyway, is there a penthouse on the roof, is there a garage behind the main house, etc.


So Google Street View is chopped liver? Flat gray 3D representations don't seem to answer those questions as well as just tooling around the building in Google Street View.

Google's also moving towards adding 3D topography to their maps so I don't quite see what your point is. This isn't an Apple vs Google thing; they're both doing the same things, they just seem to have different priorities about the order in which they do it.

> So Google Street View is chopped liver

Yes.

Because this is missing an important point. In order to render the data, the data has to be in a format understood by computers, which street view is not.


Its great especially as a duckDuckGo user. I no longer have to use Google/Google Maps to see a satellite view.

Better, more accurate maps are a good incentive to keep people using their platform. It's an incentive to continue using iOS, as now even better maps are built it! (This is my quick off the top logical explanation.)

potential bet on self-driving cars in the future - you'd want higher fidelity maps for those.

Also, they can get more accurate navigation for pedestrians etc.

Overall, IMHO, this is an investment in user experience.


I don't think the fidelity here is high enough for self driving vehicles, but the expertise built-up in the process would pave way for next steps.

Agreed, but also we don't know what exactly Apple (and also Google) is collecting - Apple Maps might only have subset of their data.

Not to mention, having a good way to build realistic virtual environment for acceptance testing is probably key here to enable self-driving.


Privacy and no ads are also features/competitive advantages. Also avoiding being beholden to Google.

They're chasing parity with Google Maps

At 4:45p PDT, this post was #14 on the HN front page with 77 points and then it suddenly disappeared.

I can't find it on any of the first several pages of HN. Where'd it go? Was it flagged?


Between this improved map and Look Around, reasons to bother with Google Maps are dwindling.

Using Yelp as a POI provider prevents me from completely dropping Google Maps. Even ignoring Yelp’s shitty practices, needing the Yelp app to view reviews and images is a showstopper. Apple needs to dump Yelp and go full Google here their own reviews and POI DB.

I don't believe you need the Yelp app. It will prompt you to use it, but reviews (and seemingly all other information?) can be viewed without the app.

Edit: never mind. I’m wrong. I could have sworn I used to be able to just view the web version of yelp last time this came up, but I might be misremembering the browser vs the Maps app experience.


Tested on iOS 13: I can scroll through thumbnail previews of images but tapping to view full size opens the App Store to Yelp. As for reviews, it shows 3 of them as shortened previews, so I can’t even view a single full review (unless it’s trivially short). Tapping on them again opens the App Store to the Yelp app. Around the review previews are links to “Open Yelp” and “Find out more on Yelp”, all leading to the same thing: App Store.

This is such an un-Apple-like UX. Idk if they signed a long-running contract with Yelp out of desperation when Apple Maps launched forcing it to be this way, but if not, then their insistence on this experience is truly dumbfounding.

Edit: I see you’ve tested it yourself.


My two cents, Apple Car edition: adding such detail increases the visual noise when glancing at the navigation display, where simplicity should be king when the user is driving.

The fact these structures are visible while browsing the map does not require that they are shown during navigation.

Honestly I never saw a point of adding these 3d structures on a map. Map should be 2d, these grey blocks don't really add much information on 99% of the maps and seems like a wasted time to me, I would rather they focus more on small and hard to find roads, that's the main reason of a map

> that's the main reason of a map

That's what you might think is the main reason of a map. There are many various reasons to want to use a map. I assume one of the main uses for most of Apple's customers, and probably even Google's given they're also working towards 3D topography, is to find particular buildings, particularly businesses, or landmarks.

Don't make the mistake of thinking that your use case is everybody else's.


3D structures are great when you’re walking around a city a trying to find a specific building, because you can now see what it looks like.

It seems plausible that the 3D vector structures are there to provide a skeleton onto which the 3D satellite images are mapped. Whether you think the latter is useful is another matter...personally I think it is helpful as a way to visualize the general appearance of an area beyond the narrow perspective of a Street View/Look Around car.

I suspect Google could be using similar data for their AR maps. It wouldn't surprise me if Apple isn't planning to "invent" that also in the near future, it completely changes the reliability of walking directions.



Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: