Google Maps just fucking SUCKS now

[CtrlAltDel]

I've noticed that now in some cities (so far I've only seen it in New York City, San Francisco, and London) the roads have much more detail when you zoom in. The width and shape of the roads is more accurate, and it shows exactly where crosswalks and medians are.

That's one of the newly-announced features, indeed…but I don't see it yet on my end. Can you share a screenshot or two?

Here you go:


And while we’re at it, here’s a picture with some of those unique markers:

[empirestate]

I've noticed that now in some cities (so far I've only seen it in New York City, San Francisco, and London) the roads have much more detail when you zoom in. The width and shape of the roads is more accurate, and it shows exactly where crosswalks and medians are.

That's one of the newly-announced features, indeed...but I don't see it yet on my end. Can you share a screenshot or two?

Here you go:


And while we're at it, here's a picture with some of those unique markers:

Yeah, I just got it to appear by using a different browser. Odd that Google's own is the one that doesn't show the upgrade.

[vdeane]

I don't understand how tech companies roll out new features.  It seems like they intentionally drag it out as long as possible.  It just appeared to me today on Chrome.  As for the monuments, I don't mind the symbology, but the size is a bit large.

Overall, in my view, a legend is important to maps for reasons of clarity, thoroughness, and consistency, and my responses to your comments are largely aimed at pointing out that you ignore or otherwise dismiss the ambiguities, and omissions, and aberrations that arise here and there with Google Maps. In seems then that your understanding of what makes a good map tolerates more uncertainty than mine does. With that in mind, I don't think there's any reason to go through your points one by one, since I would largely be repeating what I've already said.

Keep in mind that Google doesn't even have a set standard for what roads are yellow and what roads aren't.  It's arbitrary and inconsistent.  Also keep in mind that the closest equivalent to Google Maps is not Rand McNally, but GPS.  Google certainly thinks of their service as being a web version of the latter, not the former.  And it's arguably always been that way, even if the symbology used to be more map-like.  MapQuest was arguably inspired by AAA TripTiks.  In the old days, it wasn't even practical to browse the map willy-nilly.  Being able to pan around (or zoom in/out with the mouse wheel) didn't even exist prior to Google Maps.  You had to click an arrow on the edge to have the map move a set distance.

Honestly, I'm not even sure how many people are looking closely enough at the maps to take notice of symbology that isn't intuitive (like the toll road markings, or reservations).  In a way, we're lucky Google is giving us what they do; they clearly see themselves as a company providing GPS directions and ways to search for business, not look at maps.  Tourist maps have traditionally not differentiated types of roads at all.

[NJRoadfan]

15 years and they still haven't added county boundaries to Google Maps.

[jakeroot]

Overall, in my view, a legend is important to maps for reasons of clarity, thoroughness, and consistency, and my responses to your comments are largely aimed at pointing out that you ignore or otherwise dismiss the ambiguities, and omissions, and aberrations that arise here and there with Google Maps. In seems then that your understanding of what makes a good map tolerates more uncertainty than mine does. With that in mind, I don't think there's any reason to go through your points one by one, since I would largely be repeating what I've already said.

I tend to ignore or otherwise dismiss claims about ambiguity because the complaints are largely rooted in the ridiculous assertion that Google Maps somehow needs to be everything to everyone. Any Google Maps design that needs a legend fails in its symbology and user-friendliness. Companies like Google, that pride themselves on strong UX design, would never let that happen.

[I-55]

15 years and they still haven't added county boundaries to Google Maps.

It specially annoys me that if I search a county, it shows the boundary, but the moment I zoom in to where I can make out buildings the line goes away.

[CtrlAltDel]

Companies like Google, that pride themselves on strong UX design, would never let that happen.

Again, I disagree. Google drops "strong UX design" whenever it finds it can make a buck by doing so, like with the YouTube app, which used to continue playing videos after it closed, but now doesn't . . . unless you pay.


Honestly, I never thought anyone would be so dead set against having a legend accompany a map. I mean, it's not like Google would force you to look at the legend before you could use the map, or require you to take a quiz. It would be a small unobtrusive button that you could click on if you were confused. It would take up less space than the three — count 'em, three — icons they use for the photo bar:



And it's not like a legend is some weird esoteric addon that only exists because Google is driven to be everything to everyone. It's a standard element of maps, of many sorts, and has been for a good long time.

[Rothman]

I have to agree that arguing for no legend is a strange hill to die on.

Regarding vdeane's point about their inconsistent symbology, they should fix that and then provide a legend -- even if it's just a simple reference page somewhere.  Doesn't have to be right there on the map.

[jakeroot]

I have to agree that arguing for no legend is a strange hill to die on.

It's not at all. Not providing a legend is not unusual in mapmaking. It's entirely dependent on the map itself and what symbology is otherwise used to depict map elements. If the elements and symbology are straightforward enough, there is no need to provide a legend.

And it's not like a legend is some weird esoteric addon that only exists because Google is driven to be everything to everyone. It's a standard element of maps, of many sorts, and has been for a good long time.

It's not as standard as you make it out to be. Map legends have been useful in the past because static maps do not have the ability to be manipulated on the fly by the user; I don't need Google Maps to tell me that this symbol against this color is a park because I can zoom in and clearly make out where it says it's a park.

The only two exceptions to this rule are:

(1) base map colors: Google Maps uses a few different colors, but the colors are based on real-life land cover (white for glacier/snow, beige for desert, green for forest or grassy areas) and should be easily interpreted by the user;
(2) road colors: there are only two anyway, with the main difference being thickness, layering, and at what point they become visible from different zoom levels.

Have your legend, fine. But I don't know what is supposed to populate that legend that isn't already clear enough from browsing the map itself.

You should draw up a quick concept to accompany your argument, or perhaps link to another digital mapping service that does have a legend as an example that Google could follow. OSM has a legend, as I've mentioned, but it does not show more than two road types (motorway and main roads) and there is no mention of the base map (OSM does not use "geographically-correct" base map colors like Google Maps).

[kphoger]

So we should create legend to label things that most users already understand? Legends are not a requirement for a map. I don't need a legend to tell me "beige = desert".

Yes, they should create a legend to label things that most users already understand.

The reason you understand that beige=desert is because you have the experience of having looked at other maps with a similar coloring system and a legend to tell you what they mean.

Show my sons a digital Google map, and they won't possess that same experience with paper legend-containing maps.

[jakeroot]

The reason you understand that beige=desert is because

...I'm a living, breathing human that also knows grass is green, and water is blue, and glaciers are white. I don't need a legend to tell me something obvious.

The colors of the environment are obvious to anyone who has ever looked outside a window. Those that haven't likely aren't Google Maps' target audience.

[hotdogPi]

The reason you understand that beige=desert is because

...I'm a living, breathing human that also knows grass is green, and water is blue, and glaciers are white.

And protected areas are darker green than normal undeveloped land, and major roads are orange, and airports are light gray, and railroads are dark gray, and pedestrian paths are dark green, and hospitals are red, and shopping malls are orange, and state lines have a literal dashed line running through them...

[jakeroot]

Here are two maps.

I want to be fair here: the map on the left (or top) 100% needs a legend. There are so many different colors there, I couldn't possibly understand them all from intuition alone. On the right (or bottom), the fairly-simple Google Maps really doesn't need a legend. There's only a few colors being used. If there was a need for a legend, it might be for the symbols for POIs, but not the colors.

[jakeroot]

The reason you understand that beige=desert is because

...I'm a living, breathing human that also knows grass is green, and water is blue, and glaciers are white.

And protected areas are darker green than normal undeveloped land, and major roads are orange, and airports are light gray, and railroads are dark gray, and pedestrian paths are dark green, and hospitals are red, and shopping malls are orange, and state lines have a literal dashed line running through them...

But most of those things are otherwise labelled. The original complaint was related to unlabeled colors like deserts or forests.

edit: legends are primarily used to label things that cannot always be labelled on the map itself. If you have an irregular shape on the map, in multiple places, it may not be possible to label that shape on the map itself, so you use a legend, repeat the pattern and then label it so users know what it means. But if that shape is large enough to be labelled directly, there really is no reason to also duplicate it on a legend.

[US 89]

edit: legends are primarily used to label things that cannot always be labelled on the map itself. If you have an irregular shape on the map, in multiple places, it may not be possible to label that shape on the map itself, so you use a legend, repeat the pattern and then label it so users know what it means. But if that shape is large enough to be labelled directly, there really is no reason to also duplicate it on a legend.

Disagree. Say, for example, I'm looking at a map with a few areas of green with the label "X National Park". There is another unlabeled green area - should I assume it's another national park? That would be the guess, except it turns out that was a National Recreation Area because nobody bothered to tell me green applied to all protected areas.

What Google has done lately that really bothers me is that such areas are NOT necessarily the same color...but only based on what kind of vegetation they think there is, not category. The various shades of brown or green change to a slightly different brown or green and it's really not obvious that this is supposed to represent a protected-area boundary. Meanwhile, Google Maps is useless to find where, say, Capitol Reef National Park is since it's impossible to tell it apart from neighboring national monuments, forests, and recreation areas.

If I look at northern Utah on Google, I see at least four shades of green. There's a general one that seems to refer to forested areas, but it also could just mean grass, seeing as it's used in some farming areas (how am I supposed to know?). Another appears to be that same category but in protected, and another but in Indian reservations. There is a fourth, darker greenish gray color that I have no idea what it is supposed to refer to. Likewise, how am I supposed to know the large gray area west of the Great Salt Lake refers to military land if I don't know that already? And I'm curious exactly what it refers to anyway, since there are a few separate areas of military land out there and I'm pretty sure their boundaries are a bit bigger than what Google shows.

[Scott5114]

That's not great symbology–without being told I would have never guessed it was a tribal reservation, and wasn't even sure it was meant to be a polygonal area and not, like, a few prominent railroads or something that intersected.

If you zoom in another click, the name pops up in the traditional all-caps style used to label things like this:

I wouldn't necessarily associate that name with the border since there's no visual association between the two.

But, we also discussed how Google Maps is, in many ways, intentionally different from those earlier services (like paper maps). For example, you really don't need a legend because, well, (A) there's too much on the map to have a reasonably-compact legend, and (B) due to 'A', you can simply click on the features you want to know more about.

"It's a bad design, but it's intentionally bad, so it's okay!"


Just as an example of how the best-intentioned "intuitive" map designs can fail, check out this portion of my group's Dungeons & Dragons slippy map, designed by yours truly:


The legend for this map basically apes the 2005—2006 KDOT map, which is one of my favorite maps of all time, and as such the background shading is meant to convey relief/approximate elevation (although I don't have the elevations nailed down in feet in the fictional setting, because it's not particularly relevant to the game mechanics):


I had thought that shaded relief maps were commonplace enough that I didn't need to explain that particular feature of the map to anyone. I also figured it would be pretty obvious as you scrolled around that the colors indicated elevation because the features in the northeast corner of the map are mountains (and are labeled as such at higher zoom levels).

Our DM (the person running the game) sent us on a mission to the town of Megantic (in about the center of the map snippet shown here), where we were to meet up with a dragon and obtain needed medical supplies from her. We get to the dragon, and it turns out she's a brass dragon...which is a desert species. The DM had, for months, been interpreting the shading as indicating vegetation levels, and thus the tan/light orange colors of the foothills of the mountains as desert! Then I felt the need to explain to her the elevation shading (and we just patched the monster manual so that in our game brass dragons are mountain-dwellers instead of desert-dwellers) ...but I'm still getting questions about the shading two years later.

Obviously a very low stakes and silly example, but never assume that an "intuitive" map legend will be interpreted correctly–because it won't!

[jakeroot]

I really don't have the time to argue about this anymore. So let me summarize the situation:

- I have asked at least a couple times for someone to draw up a conceptual legend for Google Maps. Nothing so far.

- I have asked for any examples of another digital mapping service legend that Google Maps could base a legend design on. Again, nothing so far.

What I am contuining to see is people here picking out specific situations where legends might be helpful. Maybe so! But then, I would continue to argue that improved symbology is the solution, not giving up and simply putting in a legend. For example, I do not like how they've designed the tribal nations. I would not put these on the map as it muddles the colors. I would prefer parks and forested/grass areas contrasted more strongly. Maybe a couple more things. But I would continue to advise against a legend since I have seen virtually no precedent of a legend in digital maps as, by and large, digital maps and paper maps are quite distantly related.

[Rothman]

I really don't have the time to argue about this anymore. So let me summarize the situation:

- I have asked at least a couple times for someone to draw up a conceptual legend for Google Maps. Nothing so far.

- I have asked for any examples of another digital mapping service legend that Google Maps could base a legend design on. Again, nothing so far.

Given that no one is under obligation to provide either and neither is really pertinent to supporting either side of the argument, how you portray this as some sort of victory is illogical.

[jakeroot]

Given that no one is under to provide either and neither is really pertinent to supporting either side of the argument, how you portray this as some sort of victory is illogical.

Yeah, because precedent is totally illogical.

Google Maps introducing a legend would be a first for "consumer-level" mapping. Microsoft, MapQuest, Apple...no one else has a legend either. Do they all need it too?

I wanted a concept because I don't know what would be on the legend. I just wanted some examples of what people think it could look like.

[Rothman]

Given that no one is under to provide either and neither is really pertinent to supporting either side of the argument, how you portray this as some sort of victory is illogical.

Yeah, because precedent is totally illogical.

Google Maps introducing a legend would be a first for "consumer-level" mapping. Microsoft, MapQuest, Apple...no one else has a legend either. Do they all need it too?

I wanted a concept because I don't know what would be on the legend. I just wanted some examples of what people think it could look like.

The thread is about Google Maps...

[jakeroot]

The thread is about Google Maps...

Same tired argument. Yawn.

[Rothman]

The thread is about Google Maps...

Same tired argument. Yawn.

Same nonsense?

[US 89]

Mapquest definitely used to have a legend.

[jakeroot]

The thread is about Google Maps...

Same tired argument. Yawn.

Same nonsense?

This thread is a memorial to Google's constant ability to set new lows, yet when asked to provide an example of a mapping service with proper symbology and legends...radio silence. Precedent is important.

Mapquest definitely used to have a legend.

Maybe. Still, they're part of the overall trend that digital maps do not have legends. OSM is really the only holdout, although for good reason as their maps are far more complex.

[Rothman]

The thread is about Google Maps...

Same tired argument. Yawn.

Same nonsense?

This thread is a memorial to Google's constant ability to set new lows, yet when asked to provide an example of a mapping service with proper symbology and legends...radio silence. Precedent is important.

Precedent?  So we can't do something that may have not been before?  And yet, as has been pointed out, MapQuest had a legend as Google's main rival back in the day?  I might still be using MapQuest if their vector set didn't go to pot...

It is quite legitimate to point out that a lack of legend is part of the reason why Google Maps suck. 

And, I also get annoyed by just blocks of green where you can't tell where one park ends and another begins...

Google Maps just could be better.