What Happened to Google Maps?
justinobeirne.com
justinobeirne.com
That said, I also think the focus on paper maps is misplaced. Old-style road maps had to answer the question, "how do I get there from here?" New-style digital maps don't need to answer that question any more! Questions new-style maps need to answer include:
* I know the name of a place or street; where should I zoom in to see more things around that place or street?
* I need to go to a (gas station / rest stop / hospital); where's the closest one?
* How would I get home from where I am?
* I'm in an unfamiliar place and would like to go "downtown" (where there are restaurants and things to do); where is "downtown"?
* Where is my car right now?
Roads help you to orient yourself with the map, but they aren't as fundamentally important to digital maps as they were to old-style road maps. The visual space of the map might be better spent helping answer questions like these.
This kind of crowded display is fairly common now. I use Here Maps now which is much better + I like supporting a second option to avoid one service getting 2 monopolistic.
My theory of why google maps is getting worse is there must be designers dedicated to this that need to constantly look for 'new' things to add even if its to the detriment as they can hardly say its all good now and twiddle their thumbs. Maybe that's unfair but I've seen this effect on other digital products first hand.
Also, it's forking is completely fucked. If there are 2 right turns coming up it does not indicate which lane I should be in. In once case I had to drive 60kms, 30 kms there and back, because the interchange had 3 right exits and I chose wrong.
Now I use Google Maps to find the location and copy the coordinates into Apple Maps. Or use Waze.
My big complaint is that Google maps announces I'm on the fastest route even when I know I'm not!
Touch here to dismiss, touch there to set as destination. It takes like 30 seconds for it to automatically disappear. If I'm on a route I'm familiar with I just let it disappear...but times where I'm going somewhere new I frantically lean forward to my dashboard to dismiss their ad.
If they want to prevent accidents stop popping up ads that take up half my map while I'm driving down the highway.
Why is that last button needed?
I've noticed when purely following GPS directions that I can't remember how to get there, even after a few times following it. Knowing the area helps you take your own informed alternate routes, know what things are on the way that you might want to detour (oh, I needed to go to the post office anyway), recognize if the GPS is going to the wrong destination because of user error, etc.
In their perfect world you don't orient yourself with the map. You tell the map what you want and it tells you how to get there.
Google maps labeling is wrong and dangerous for drivers due to the need to keep your finger there.
Maybe people forgot, but google maps was /blazing fast/ in the beginning.
Nowdays, it brings my browser down to a crawl even before the images are shown. Maybe people are just stuck with this "google is the best" mentality, but this has stopped being universally true since many years.
Use OpenStreetMap. It's data is way superior. It's _your_ data. Cannot strett this enough.
Want a fancy browser? Nokia maps have always been incredibly sleek to use:
https://maps.here.com/
Google has still the lead with street view, but for the actual maps I really encourage you to look for alternatives. They've destroyed their interface as far I'm concerned.
What do you mean by "the lack of precise zooming is super annoying"? In my experience, zooming in Bing Maps and OpenStreetMap are very fast, and only in Google Maps you're waiting ten seconds for the region to load and browser to respond before you can zoom further.
To test out, try modifying the "10" in both of the links below: https://www.google.com/maps/@40.7058316,-74.2581905,10z
https://maps.here.com/?map=40.71455,-74.00714,10,normal&x=ep
Mapbox has one, Mapzen has one, various apps have their own vector map implementations, there are convertors that take OSM data and output maps in the custom vector formats used by GPS devices and so on.
And there's no way to expand it. I do understand that they don't want it to keep expanding beyond a certain point, especially as they add more details as they zoom. But they could let it expand within some constraints, such as the name of a street within the stripe of the street, and cap the size at 3x the normal size or some such thing.
If the algorithm for letting the labels grow a bit is just too hard for them, then just give me a little slider on the side to add a fixed, default label magnification factor. I could expand it a bit to make the labels easier to read in general, expand it more (but just temporarily) to get a better look at something, or shrink it a bit in cases where the labels were harder to read due to overcrowding.
As it is, they've removed too many useful labels without expanding the ones remaining to make them easier to read and prevented me from expanding them manually. Bother.
He, came here to post the same thing. Since the last 2 or 3 years I had the impression GM went worse (for what I do with it, not even mobile) after every update made to it, and sometimes also slower and more CPU hungry. To the point I stopped using it and went looking for alternatives.
Especially driving country roads and having to zoom in to kite level in order to see the medium sized villages is ridiculous.
Anyone know if Bing Maps is better?
- OpenStreetMap / OsmAnd for general use (about 75% of the time);
- Google Maps for traffic info, street view and businesses (about 20% of the time); and
- Bing Maps for areal imagery (about 5-10% of the time).
Google Maps is always extremely slow, but their business listings are very comprehensive due to their extensive marketing, and traffic info and street view are also expensive things open source cannot really rival.
Bing Maps runs at the speed of light when zooming and panning around the map. So much better than Google Maps, but the map's quality is not great so I just use it for areal imagery (another thing OSM does not have).
OpenStreetMap is as fast as Bing Maps and its map quality is better than both google and microsoft if you disregard business listings (at least in western Europe and the middle of nowhere[1]).
Landmarks are key to any map, but landmarks that fade in and out of focus are not all that useful. I used to be excited by the prospect of being able to bring up a map of my route and being able to see where I was going on my phone mounted to my dash. But as of late, I've been struggling to find the right zoom level that shows enough detail of the area I'm traveling while showing enough of my route.
On a number of occasions over the last year I've has to pull over and reorient myself on my map due to a failed pan/zoom attempt.
It's funny that this article came out at this time as I've been evaluating ways to mount a larger device (tablet) on my dash as maps on my phone has gotten to be rather cumbersome.
I feel that for the most part the details in this article are accurate, that the attempt by google to make the maps load quicker on mobile have compromised critical details available on the maps.
One of the key areas where this could be addressed is by loading details based on need. For example if I select a travel route between two locations, load more of the details related to that route and reduce the extras that fall outside my concern. Show me roadways that leave my target route, as well as the cities and towns along my route. Making an attempt to provide me the details I need without my need to interact with them as much as possible would be great.
What's the actual problem you're trying to solve when you do that?
Consumers of OSM data generally do perform a data extraction step where the freeform tags are grouped together and perhaps normalized to some extent. So at that point you sort of start to have layers, but different apps will use different systems and rules for that step, so the layers are coming from the app maker, not from OpenStreetMap itself.
Interactivity is a failure state of software that could have predicted what the user wants from context and past behavior, and tailored the display and presented it to them without being asked.
"Further, the user might prefer to learn information while using her hands for other purposes, such as writing or eating or stroking a cat [or driving the car]. Each time software demands the user’s hands, this activity must be interrupted [at the risk of causing an accident]. Finally, the growing prevalence of computer-related repetitive stress injuries [and using smartphones while driving] suggests that indiscriminate interactivity may be considerably harmful in a literal, physical sense."
That is not true. A good paper (road) map has three zoom levels. 1) A major map covering most of one side of the paper, with enough information to get you easily across the focus area of the map. 2) Often on the side a smaller, sparser macro area map, showing only the freeways and the major cities/towns, often spanning a slight larger area than the main map. 3) A few downtown/focus maps, that are far more detailed than the main map, showing more rodes (sometimes walking paths, and bike lanes, house outlines, etc.). Often on the side along with the macro level map, or on the back next to the index. And 4) a really good paper map has also a macro level public transportation map (often on the back as well) at a zoom level between (1) and (2).
The complaints of the OP are perfectly valid but they're more a matter of taste than an example of bad design decisions.
Maybe the OP would be satisfied with a couple of sliders?
Less Streets-------------|--More Streets
Less Cities-------------|--More CitiesThe problem I have Google maps lately, is that i have to zoom so far in to see the names of some roads. To the point where nothing but the road is on screen before the name appears. It's a major UX issue.
That said, I suspect that there's a fundamental difference of processing maps among people; I like to store a map in my head and trace mentally. When I drive, I prefer to see the map always north up, so it's easy to synchronize with my mental map. However, most car navigation systems orient the map as your direction---I find it very uncomfortable, but apparently that's most people want.
Doesn't seem too bad.
Of course, you DO have SOME idea of what you're looking for and that's where Google maps can help infinitely better than any paper map.
I use OpenStreetMap for this use case so I can only answer for them, and they seem to do a great job at it. If I am looking for the general shopping area, or office areas, or anything like that, I can zoom and pan through a city and find it very quickly. That Google doesn't show labels anymore makes it completely useless (and I've run into that a few times).
"Ok Google, show me a map of Oakland, California."
1) Answer "Yes."
2) Show a route to San Jose.
3) Play the song.
You can't zoom out if the information is only showed on a tight zoom. You need to keep zooming in and out.
Here's an example of how tight the zoom has to be to get eg bus stops to show up. For this example the user has to zoom in really close to get the bus stop shown, then either very careful start zooming out while keeping an eye on where their bus stop was OR keep panning.
So, if you plan a route, do the maps change depending on your destination?
Google, hire cartographers. Amazon, hire librarians and typesetters. Spotify, hire musicologists.
I.e. when you look at a Google map it is using the latest GIS data to construct a “good” representation for your screen resolution using your specified zoom factor and possibly even with dynamic overlays or highlights.
Likewise, when you read an e-book you can read it on many different display sizes, pick your own font, and font size. So where we used to have the book typeset once by a person, it is now done on-the-fly using user and device specific parameters.
The result won’t be as good as when you have a human do it, and even when a human does it, there might be limitations in the e-book markup language, just like with HTML.
Strongly disagree with that. For one thing, Spotify already has an army of tastemakers spending all day assembling curated playlists. For another, they have the Echonest data which relies heavily on manual labeling.
But the major innovation in Discover Weekly was to use machine learning directly on mel spectrograms to figure out meaningful features for human taste. They still want to rely on their experts as much as possible (and hey, I don't blame them for not wanting to fire people), so they try to combine their expert's features with the algorithm's. But this introduces human biases again.
The problem is that when it comes to music, everyone's a missionary. Everyone wants the world to listen to the music they are excited about. Professional opinion-haver about music is the dream job for many adults, much like chocolate factory taste QA expert is for 6-year olds. And they just can't separate their own opinions from objective truths very well. It's hard to be objective about something you love.
The real great thing about AI in recommendations isn't really the intelligence part. It's the "AO" - artificial objectivity. The algorithm is probably inferior to humans in some aspects (it can't interpret the themes in lyrics very well, for instance), but the advantage is that you have full confidence about
1. What information it actually might use, and
2. What it tries to optimize.
From point one, you can be sure that it's opinion on Smashing Pumpkins isn't affected by that annoying kid in 8th grade that used to listen to them. For point two, you can be sure it's really trying to find the music you will love, not what it thinks you should love.
To get it slightly back on track: I can't wait until an AI can do music history, or etymology, or economics, or history. Or matchmaking in dating! It will be useful long before it can match humans on intelligence. How great wouldn't it be to get results in those fields which you could trust were from a disinterested party.
A more important task would be to restructure the "information architecture," for example to improve the experience of looking for classical music or jazz. There's a lot to do that isn't just based on opinions.
The dream of AI doing music history seems kind of bizarre to me... as well as the whole idea that human knowledge is bad because it's biased...
To the degree that judging good jazz or classical is different from judging other types of music, I think that yes, it's based a lot on opinion. In particular the opinion of authorities - critics and other performers.
It's not that this is entirely unreasonable. With music as a social phenomenon, you might prefer to not be "into" the wrong kind of music, even if you would like it for the music itself. Spotify and Echonest have actually talked a bit about how listening patterns can reveal "shameful" tastes, different from the tastes we would like to project.
The job of a recommendation system then, if we should look cynically at it, is to show you the "right" kinds of music that you would like to like, but actually like too - and to not tempt you with "wrong" kinds of music that you would like despite yourself.
And yes, I'd bet you'd need musicologists (or human analysts) for that. It would by definition be very hard to figure out from listening patterns or acoustic features. But isn't this a bit cynical as I said? Shouldn't we try to not be ashamed of what we actually like?
Categorizing, labelling, organizing, displaying music information is not about judging quality.
That aside, I also don't believe it's possible to separate a "pure taste" from a "cultural taste," philosophically.
My hunch is that the tech industry is bad at taking advantage of this kind of traditional trade knowledge, and the essay in question here is an indicator.
As for Amazon, the Kindle e-books very often have horrific typesetting that would be laughed out of any traditional book printing shop.
And as for Spotify, their catalogue mechanisms and information design are clearly awful from a musicology standpoint.
It's an example of an industry and body of knowledge that still grows and is passed around outside of the internet.
Weather patterns for a particular area, places to anchor and dock, safe passages through reefs. You can't always google the answer and much of it is gleaned from other sailors who used experience and their tools to figure it out.
Having grown up pretty much constantly online, with the answer to any question I have being reasonably available with a quick search; that some common information about the world can still be discovered and shared is is really fascinating.
It sometimes feels like there isn't much outside of STEM fields left to explore.
I've been online and windsurfing for decades and I've watched online info about weather/wave conditions at various locations grow then peak, and now it's been declining and drying up in recent years.
It used to be curated and published on websites and talked about in public discussion forums. Even usenet and mailing list archives dating back to the early 90s would appear in searches. In the last 5 yrs or so it's all been slowly evaporating into ephemeral Facebook group posts hidden behind a login form.
I suspect this is happening to all sorts of other communities too as they drop below critical mass. I'm not thrilled about it.
Lots of this information is collated in Almanacks and pilot guides (eg. Reeds Natuical Almanac and the Shell Channel Pilot for UK sailors, plus the books from the Cruising Association and Royal Cruising Club). There's not a lot of incentive for an online version, as the information needs to be accesible offline for reference and in case plans change. However, there is a lot of information in apps: there's an iPad version of Reeds Natuical Almanac [1], Imray has a worldwide Tides planner app [2], and there are apps with charts from both the UKHO and Imray [3].
[1]: http://www.reedsnauticalalmanac.co.uk/
[2]: http://imray.com/tides-planner-app/
[3]: http://www.yachtingworld.com/blogs/elaine-bunting/testing-ip...
Other apps I've used that have a slick interface are unfortunately online only. Which is useless to me. When I'm coming into a new, unknown, anchorage I need the information now, not after I've already tied up.
As is the fact that other than Knuth, no-one seems to have learnt the first thing about how to lay out a page of text in an attractive fashion. And then even given Knuth, essentially no-one bothers to use his magnum opus (I honestly think that TeX is even more important the The Art of Computer Programming, and is what Knuth will be known for in two centuries).
My biggest gripe is contrast, rather the lack thereof. Zooming in and out doesn't help. There's a lack of contrast at all levels!
And the algorithm for displaying place names sucks. You'll see certain names at one level, zoom in and they disappear, zoom in some more and they finally reappear.
Paper maps are unquestionably more ergonomic (but much less convenient) than Google maps. But it's not just Google. I find other online maps equally bad. It's quite sad that a paper Rand McNally map is so much better at actually presenting the geography of an area.
Perhaps other posters here are right, it seems like Google maps is designed for point-to-point navigation, nothing more.
On the whole the HTML and CSS on that blog are a mess though.
Firefox's Reader View helps here by bypassing the author's styling completely.
Google provide an alternative mapping product, Google Earth, for satisfying curiosity about the planet. Google maps is primarily a navigation tool. They have very distinct use cases.
I'm not sure that puts him in a position to be trusted to criticize google maps. Apple maps has been considered inferior to google maps since its launch.
Just because he has ideas about how to pack an overwhelming amount of information onto a map doesn't mean it's a good idea for normal users.
"I designed and led the development of Apple’s cartography, and I founded, recruited, and managed Apple’s multinational Cartography team."
For navigation there are separate websites that use OSM data, like yournavigation.org (not sure that's the best one, I just know of it). They often work a bit differently than Google Maps so there might be a slight learning curve.
There is also Bing Maps which has great areal imagery if you're looking for that, only their map quality isn't as good. Certainly not terrible either, though.
You don't get customer feedback like this (https://productforums.google.com/forum/#!topic/maps/HfC6dYi4...) when you're doing things right.
Old style printed maps had cities on them, because the map didn't know where you were going! You had to find your city or your location on the map. Now the map knows where you're going, so it can show that place extra-clearly while hiding a lot of detail that's not relevant. Roads are relatively more relevant than cities, since you travel along them to get from one place to another: displaying a road shows the user that they have a primary thoroughfare between locations. You might not care about the name of a city if you're just passing through; and the city that is your destination will be specifically shown.
My guess is that they display only as many cities as needed to help people orient themselves while looking at the map, to understand what they're seeing. More than that is irrelevant to the primary use-case of navigation.
> Google Maps of 2016 has a surplus of roads — but not enough cities. It's also out of balance. So what is the ideal? Balance.
The ideal is utility, and the key use-case for Google Maps at that zoom level is driving navigation. The user's going to input their own destination into maps anyway, most of the time, and they'll expect it to appear, so it's no surprise when it does.
Google would have data on this: how many users use Google Maps while driving regularly, multiple times on a trip (at that zoom level), while not having a destination entered (and with no destination, obviously no turn-by-turn directions)? Probably not many. Now imagine overlaying your route with current position and destination on the maps - it's going to be easier to scan the new ones. Edit: Navigation is the primary use-case for a map, and I'd guess usage motivated by that purpose dwarfs the rest by an order of magnitude, and so it's a good default.
I personally have found the new google maps very frustrating and slow - I'm constantly zooming in and out. This article finally made me realize what I'm missing, and why I'm having so much trouble with it.
The place where Google maps really excel is applications that combine a search result of some kind with geographic data. For ordinary navigation, especially if you want geographical context rather than just directions from point A to B, traditional maps still are much better in my opinion.
The new data that Google has comes from Android handsets and from users using Google maps and Waze on Android and non-Android handsets.
This data is all about users in motion. At the scale shown in this article, it's almost exclusively people driving. As a result, it makes a lot more sense to focus on the connections over the places they connect. This becomes clearer when you view the roads as more active entities by including congestion and other real-time data.
This may not be the best presentation for everyone, but it seems to be the presentation that fits best with Google's current mission and capabilities.
The paper map was used to navigate from place to place. That never happens with a google map. Sure, you navigate with them, but by telling Google where you want to go and letting them draw a line on your map. You don't need all that extra information if your phone is navigating you from place to place. You just need something clean that you can glance at to get a sense of where you are.
So that's what they've designed their maps to give you.
https://www.openstreetmap.org/#map=9/40.5263/-73.8556
https://www.google.fr/maps/@40.5487361,-73.9399427,9z
(Edit: as the child comments correctly point out, this is only the default OSM rendering style.)
There's even several other example renderings that can be easily accessed on the website, for example
https://www.openstreetmap.org/#map=9/40.5253/-74.2003&layers...
"ugh that is so ugly look at all this clutter"
ps: I recently discovered the 'my timelime' feature. Surprising to say the least.
At the zoom level the screenshots are taken at, maps are essentially useless. The most important information they can convey is "there's lots of roads here" or "this region is densely populated". The maps aren't optimized for accuracy, they're displaying a summary. The long island example really struck me - the old map displayed the primary route only, the new map conveys the fact that there are multiple options. If you're stuck in traffic and you pull up the map, you can see there's another decent route and ask the app to provide you with directions on an alternate route. If you're using the old maps, you'd just see the single primary route highlightedand assume you should stick with the route you're on.
Default rendering: http://i.imgur.com/uhbRcsR.png
Find alternatives, there are several mentioned in these comments for starters.
When/if Google start seeing a reduction in their map use, only then will they start paying attention.
Not applicable at the zoom scale used in the article, but on higher zoom levels this is certainly a factor.
Another thing I'd like to see is making the "avoid tolls" setting easier to get to. Northern Illinois is toll road central, and I-355 in particular is a huge ripoff when you pay cash. Since I don't need any of the tollways for commuting, I can't justify getting an I-Pass.
It doesn't even remember the setting, you have to hit the three dots for menu before you search, choose "Route Options", then tick "Avoid tolls", then hit OK. EVERY TIME.
And if you forget to go through this and start navigation, there's no way to change it, you have to hit back, then do the above. Why not have a way to change route options from the navigation screen??
https://en.wikipedia.org/wiki/Cartographic_generalization#Th...
On google maps I can never find what I want. I thought it was because I've been using OpenStreetMap, and had gotten used to a different display style. Seeing a place once in OSM anywhere in the world, and zooming out from it to continent view, I can almost always find it again later. On Google Maps I always got lost. Now I finally get what the problem is.
That or people just find whitespace aesthetically pleasing and Google designers went kind of crazy with it.
So I'm not too bothered by this change. What I don't understand is why hasn't anybody taken the Google Maps routing and use it in OSM apps? Might not be legal, but similar non-commercial projects it should be fine.
(Small example of what I mean: many people like the new style because it's prettier, but when I use OSM I get to places a lot faster than people using Google Maps, so they might be asking the wrong question even if they did user testing with a million people. But that's just one example of everything that could go wrong.)
Which just makes the changes to maps so baffling.
Anecdatum: last month I was driving out for a weekend in some rural bungalows a few miles outside a small city (Elvas, Portugal). The address was a bit vague, the place name too common, so much so that I had GPS coords stowed away in a message pic (don't ask ;-).
So, when I pop out Google Maps in the old faithful iPad2 (which happens to be the 3G version, and therefore GPS chipped, good for navigation), and zoom in the area ... amidst thin local roads and lots of blank space, there's the place name and the days we're staying there.
Turns out the Gmail app in that iPad was also used to send or review the emails with the reservations.
Even in my 'desktop' I've been noticing Google Maps marking out city places which seem small compared to other landmarks, but where I often go or mention in emails.
(Thanks, I do know where I left my keys today, I'm good.)
Off-topic, but as a native NYCer, we would never call it that. It's the LIE. I once had a woman ask me in the parking lot of a Walgreens how to get on the 278. I was puzzled for a second, then I realized she was talking about the BQE. Living out in California now, I miss the days of calling highways by name.
You don't need a hulking great map with loads of detail at a high level to get from point A to point B, you now just use your smartphone for that.
I assume Google spotted a trend of people searching place names as opposed to picking points between two separate locations.
So how does that change the function of the map?
Well, we no longer need to have the zoomed out overload of detail, if we need more information about a place we are visiting, we type in the city name, or address, then zoom in close to see the detail we need.
The article kind of skimps over the point that we can interact with those maps now.
If you need that fine-grain detail zoom in?
Not saying I advocate this approach, but thought I would try give you an answer :)
Not sure I agree, in my opinion most of us use search & destinations nowadays, even in offline mode. The only reason I look at a map is to gauge distance between me and my destination.
I certainly do
And for this use case Google maps is getting pretty useless
Offline usage is best for travelling.
Google’s products are shit outside of the US, more at 11
It’s nothing new, it’s been like this for quite a while.
No designing human ever sees a rendered Google Map, except for testing. The overall esthetic programmed into the algorithms are designed, but no human looks at the final result and says "it would just look better if this was shifted up and to the right a bit." There are an infinite number of possible renderings on a digital map, but a specific published paper map goes out the door with a sigh of satisfaction.
Which is to say, there's a long way to go before a digitial map is beautiful, and a specific rendering is as good as possible for the person viewing it for a specific purpose.
The only way digital maps win (and they do) is that you can ask them things, and they cover the world.
This is a good thing because in a place like tokyo there is just too much stuff - if I care for aometgubd I either search for it, or drop a pub and check what is around.
On the other hand, Maps always shows the labels of things I have searched for in the past, yielding a customized legend of landmarks+things that matter to me.
He poses the question of which map you'd want when lost. A mobile phone with Google Maps is clearly the right answer.
It's like claiming that the new york times should display the entire full front page of the newspaper on a mobile device so you can read several articles without scrolling or loading more content because that's what you used to be able to do with the real paper.
There are many lessons that digital maps can learn from hand-crafted maps.
Justin O'Beirne knows what he is talking about. He has worked for Apple Maps and he also created other interesting and really precise analysis in the past.
http://www.justinobeirne.com/essay/google-maps-and-label-rea...
The paper map has to have lots of cities on it, because there's no other way to find where exactly the specific suburb is if it's not on the map. In Google Maps, you can zoom in, you can search, you can have a link for direction sent to you, ...
The author may claim "less is just less", but apart from "printing a map before knowing where I'm going", I can't think of a situation where the "improved" map would be at all useful.