I'll do a shameless advertisement of our open source routing engine GraphHopper as one of our main advantage over the big companies is that we do not track users. We rent our servers (not on AWS or GC :) ) and remove logs after max. 5 weeks https://graphhopper.com/maps/ (see our privacy policy for more details)
We only keep them for debugging purposes and attack prevention. The maps layer (visualization) is done from an external company though: you can change this in the upper right corner to e.g. the original OpenStreetMap or Thunderforest.
Additionally you get nice biking and hiking routes all with elevation, multiple other vehicle types like scooter or truck. And the possibility to export to GPX or using multiple via points. Additionally we try to update the road network as fast as possible (currently max. 1-3 days) from OpenStreetMap data, which is not always the case from other providers. Currently we try to revamp our UI to integrate new features like our alternative routes (you can trigger this via algorithm=alternative_route URL parameter).
Now it is always important to ask the question "how do they make money?" and to understand why we can offer this all for free and even have the routing engine as open source: the simple reason is that we do not have to earn money with the end user. We created the company to push our open source projects (we also have jsprit - a toolkit for solving rich vehicle routing problems) and we try to support OpenStreetMap where we can (e.g. we are the only European corporate gold member of the OSM foundation) and do this via our growing SaaS where we sell subscriptions to many small and medium sized businesses. We do not have VC and so there is no pressure into things we don't want.
I have only two things to say in response to this:
a) I use GraphHopper with OpenStreetMap.org sometimes. It is good. Sometimes too good. Can I ask it to stick to bigger roads at the cost of other things like travel time?
b) Thank you so much for bringing up the profit model and being transparent about it. Way too few people do that.
Do you mean preferring bigger roads for car? This is tricky as optimizing for time sometimes introduces big detours and so we consider also distance (that can take you off big roads) and so yes, we have to improve this further. At least for (small) truck we have to improve there :)
For bicycle, in my case. Sorry, should have specified. Bicycle infrastructure is awful where I live so I need to spend a lot of cognitive effort just not being killed by the road. In that mental state I can't follow very complicated directions, so I prefer bigger roads!
Bike routes are currently heavily optimized already and sometimes I'm also a bit surprised about the good route quality ;) but actually this is only possible with the fantastic OpenStreetmap data. Still the most improvements are probably possible via data improvements in many areas. Still we will try to improve routes via "less turns" and more precise elevation data etc. For bike routing we only need to keep the balance between time, distance, simple route and beautiful route ;)
Spreaking of bike routes, personally that's why I prefer Grasshopper over other route planners - exactly because it seems to show the most direct path, often avoiding large roads. Riding a bike along a road with heavy car traffic is not enjoyable.
Something you could do to be better than Google Maps would be to prefer stoplights over two way stops for left turns. That nearly got me sideswiped in SF once.
Do you have an example? (avoiding left turns is the same problem like avoiding turns in general that is possible now with some new code we have since 1.0 or so, but we have to do more work to make it working in real world)
I don't have a specific route I can link to, sorry. I was navigating in an unfamiliar area of SF once and directed to turn right one street before what turned out to be a major street, then make a left turn onto another major street. The directed left turn was from a stop sign onto a street with poor visibility and cross-traffic not stopping. If I had gone one street further before the right, there would have been a stoplight.
Basically if you have data on the type of an intersection, favoring nearby controlled intersections for difficult turns would probably be helfpul, especially in hilly areas.
Sorry for hijacking GraphHopper's thread, but: you might want to try the OsmAnd phone app—it has some settings for routes in this very vein. E.g. ‘no unpaved roads’ and such. Not sure if it would solve your specific case, and OsmAnd isn't completely free, but it's an option (specifically map updates are limited for free, at least in the Play Store version).
I have the problem with OsmAnd that, where there's a freeway offramp that briefly parallels and then rejoins the freeway, OsmAnd always chooses to take the offramp and then immediately get back on the freeway. There are no words for how incredibly stupid this is.
Is there a setting that disables this ludicrous behavior? Why is that setting set to "stupid and dangerous" by default?
I use osmand. I'm not sure whether it uses GraphHopper or not.
But, when routing from Santa Cruz to San Francisco, every time there's a freeway offramp that parallels the freeway and then rejoins it, the osmand routing wants me to take that offramp. So that I can then, just a few seconds later, get back on to the freeway.
This isn't just annoying, it's dangerous. If that's you, can you fix it? And if not, can you make sure you don't do the same thing?
I work for an IoT company doing vehicle tracking. We wanted to draw a line to show where cars had been. Google's routing algorithm insists on passing every point. GPS errors mean sometimes this resulted in routes going off a motorway, passing an (incorrect) point, and then rejoining the motorway.
Graphhopper's map matching algorithm handled this perfectly, and we're a satisfied, paying, corporate user (since March 2019).
Tech support was also extremely responsive and helpful when I needed to debug some edge cases (e.g. extremely long trips).
My friend Jessie in Germany also uses Graphhopper for visualising research on tram lines, and her experience has also been excellent.
I was not paid to say this, I'm just a satisfied customer. Use Graphhopper.
Is there any plan to add stuff like business addresses and all the other fluff of Google Maps? I'll happily use this for getting around but having to input the street address instead of just the name of a business can be a hassle. Either way, thank you for offering an alternative, this seems great.
Actually POI search is already possible and works (according to our tests and effort) in some countries like in Germany ... but geocoding is another big topic besides routing (and map visualization) that you need for a good map application. And to make it working on a global scale is much more effort as the data quality in OSM and how addressing works can be very different for the different countries.
BTW: we also provide our own geocoding which is based on photon an open source geocoding projects where we also invest, additionally to financially supporting the 'main' OpenStreetMap geocoder Nominatim project.
Sorry for the trivial question: should I leave house numbers off the start/end points? i.e. this works: <road name>, <zip>, <town name> but this does not: <house number>, <road name>, <zip>, <town name>
The article doesn't say, but if it used an OpenStreetMap-based routing engine (which you can self-host because, y'know, open data) then the directions might misalign to the display layer which is Apple Maps. Therefore it must send your directions request to Apple. Which I guess is fine, they'll just see "some DDG user" requesting certain directions and it's definitely not the worst company they could be working with, I'm just a bit surprised that DDG went with Apple Maps instead of self hosting.
It's the one thing that's kept me off DDG - the superior choice of OSM exists. And as you've pointed out, their choice of maps now forces the user into Apple's hands as the choice of navigation, a company which I do not trust, and I certainly don't see it as the least evil.
Even placing my mistrust aside, it's once again a US-centric move; as garbage as Apple Maps is, it's even worse outside the US.
The only reason for this choice that makes any sense, is that they've received a monetary incentive for this, and neither company has been open about it.
Speaking from experience, trying to deploy a production-grade OSM instance is an enormous amount of work. Several businesses have been built purely around that (Maptiler and Mapbox, to name two off the top of my head). I agree that, particularly outside the US, OSM may seem theoretically to be a better choice, but I'm more inclined to assume that DDG simply decided the development resources needed for OSM were too great to justify it, and went with the alternative that they think gives them the best shot at both improving user privacy and competing with google.
That sounds logical to me. Do you see anything the OSM community could do to make this easier? Or do we just need (more, perhaps) companies like Mapbox and recommend those rather than self hosting?
For my personal application, I needed access to the actual data, and wanted my map render to be in-sync with the data layer. For that, there aren't any companies (that I'm aware of) offering services; that would also be really tough to do, because your database data model/schema for that is always going to be application-dependent. So I think that Mapbox and co fill that particular role fairly well. But, I don't believe they offer route planning (it looks like there are other companies that do, but for DDG, their SLAs would be important), and route planning needs access to the raw data as well. So really I think the biggest impact would be in improved tooling.
If I had the time, I'd personally focus on writing a loader replacement (ie, imposm/osm2pgsql "competitor") that:
1. supported arbitrary data transforms into postgres -- basically, can load data into any database schema, do on-load data transforms, enrichment, i18n, etc.
2. was distributable via some distributed message queue (I'd personally choose SQS, but you could do something like celery for less-opinionated solution)
3. was not opinionated about deployment compute environment (ie, neither requiring nor forbidding containerization)
Basically, the ability to have a proper, production-grade, not-just-meant-for-slippymaps ETL pipeline on OSM data would be a really big deal. I haven't ever seen an open source solution for that.
But only because they can tie it to a user account. If DDG is the middleduck acting as an anonymizing layer, Apple won't be able to perform such analysis on the data.
This is all conjecture obviously, it's hard to know exactly how much data Apple is getting, and if they could use something else to fingerprint users, but I'll give DDG the benefit of doubt for now.
Navigation requests are very frequently home<->"place I went". Household-level driving/shopping patterns are incredibly valuable in their own right, but it's also easy for a dedicated company to tie your address to the personal identities of the residents (Apple's internal customer records, public records, private vendors whose whole business is selling this info), and then it's just profit all the way down.
Hypothetically, DDG could guard against Apple having useful navigational histories for your home by seeding requests with misinformation -- artificial navigation requests Apple can't tell apart from your real requests -- but I'd be surprised if this hypothetical is actually happening.
Correlate navigation requests made through DDG with location data from people with less of a focus on privacy. From there you can use a social graph to narrow down who lives at a given location to a small set of people. It's a lot better than straight up telling a third party which account lives at/uses a certain location, but it's navigation is a very revealing activity no matter what.
> like our embedded maps, powered by Apple's MapKit JS framework
The JavaScript framework they use is not necessarily the same as the server-side routing engine. One can overlay data (like the path to drive and turns to take) with any mapping library like google, leaflet, etc.
Good, I decided to finally ditch Google Maps as my usual go-to app when I need bus directions, itineraries or things like that. The app is now cluttered with too much suggestions about restaurants, places and suggestions I have no interest about on the moment.
I recently re-installed Here maps (which I loved on windows mobile) and my country railway app has built-in bus data so it can help plan travels as well.
I didn't saw any hints about it in the article but are they going to build an app around it ?
Edit: ah, they don't have public transport schedules yet.
Google is trying to find the saturation point where people start leaving...
My annoyance is that every second trip is... back home. So with their gigawatt AI, I suppose it could have learnt. So I guess it tries to do its best to avoid me going home: typing « Home » suggests « Home Depot? home improvement? Homie restaurant? Homicide drop point? » I’m joking with the last one but Google seems to be pushing too hard to make me discover new places.
I tried many times in the past but even though it's faster than Google maps and doesn't try to nudge me to death the UI is very cluttered (or more precisely: there are a lot of options and buttons everywhere). I think I also bought a license or a one year pass or something at some point.
After getting thoroughly sick of the full page "pop up" that Google Search and Youtube both now display on just about every visit, I switched to DDG for search and youtube-dl for YT and it's actually been great. DDG results are fine, and the !bang (!g, !w etc) makes it really easy to jump to other search engines / places if I really need it. Google have lost me as a regular customer for good.
I was interested in doing this too. One issue I found, however, was getting a list of my subscriptions without having to go to the website and fetch the channel IDs or video IDs for youtube-dl to then use.
I also noted that youtube-dl can download from your subscriptions list directly, which is cool. I was going to put that on a cronjob and just let it milk my subscriptions on a daily basis as new videos come out.
I've also been using "Clean Up YouTube", a Firefox plugin, which basically eliminates whatever you want from the UI. For me that's the home screen/trending crap, comments and just about everything. Now I only focus on my subscription feed and move on from YouTube after that. This has helped to drastically reduce the amount of crap I watch.
I've had constant issues getting OSMand to navigate to addresses, even after checking that the address exists on OSM. If I search for a POI, it typically has no issue finding it and routing me there, but if I enter a street address, I get nothing. Enabling online search after failing to find an address once used to work perfectly (online search uses Nominatim instead of photon IIRC), but they seem to have removed online search entirely in the last update, so now I can only navigate by POI name or cross-street.
It might be better in other places, but it's a long way from replacing google maps for me.
Indeed, OsmAnd works wonders offline on phones: there's a ton of detail compared to some proprietary services, at least in some cities, and routing works fine without internet connection (with the exception of public transport when the data is not in OSM).
It's not entirely free, namely map updates are limited without a subscription—though I haven't checked if it's the same with the F-Droid version.
Accessing maps and directions through a web search, though? That hasn't made sense for years on mobile. Sad to say, they need an independent DDG Maps app to make this viable on mobile.
Absolutely agree on this. Furthermore, there's zero interactivity with the map other than zoom and pan. Maybe it's just me, but I can't click any labeled point of interest for more information. Nor can I click on a spot to set start or end locations for directions! It's surprising how non-user friendly it is at this stage.
DuckDuckGo seems to be a US-based enterprise (remind me again how that works with being "privacy friendly"?) so it's hardly surprising as cycling seems to be almost non-existing in most cities in the US.
There is less consensus on routing for cycling, no one-size-fits-all solution, because everything depends on stamina, level of comfort cycling with traffic, tire width and wheel durability. On Android, for instance, see how different the routes can be when generated by OSMAnd's in-built cycling router and when generated by the Brouter plugin (which in turn has two profiles users must choose from). No surprise that DDG doesn't want to roll out cycle routing before they can think up a UI that offers the required customizability without daunting the user.
Graphopper can give you a bike route which avoids hills, a mountain bike route which doesn't and also
uses unpaved roads and a road bike route which also dorsn't but sticks to the road. It's quite good too. I used it to plan bike trips, because it also exports GPX tracks.
This is cool but doesn't use my most frequent use of Google Maps or Apple Maps: checking to see where traffic accidents are before leaving.
Portland has freeways running north/south (more or less) on the West side (I5) and East side (I205). Sometimes, one of them will be significantly slowed because of accidents or some other unspecified cause.
After being stuck on the freeway once or twice, it's now our habit to check the route for traffic. What's normally the fastest route can be the slowest because of traffic.
I haven't enabled DuckDuckGo to use my location, so when I go to directions, it assumes my location is in a city near me, which I assume is based on my IP address.
What's the benefit of doing this, rather than just leaving it blank?
PS: Product looks really good. Really quick on Firefox, without any dumb animations like Google Maps that make it seem so much slower.
Something I'd really like to see in all map apps is a "scenic" route option. Routes organized by level of traffic, nearness to parks, etc.
I'm usually willing to walk/bike an extra kilometer or so if it means I'll walk through a park instead of next to a highway. Unfortunately the default/only option is always the quickest one.
Awesome and thanks, DDG! No longer need to use Google for all my direction inquiries.
Now if you could please offer a front end email service for iCloud, ProtonMail and or other privacy centric email clients (let the user choose) and a DDG full fledged news service (it could rank news sites which are privacy centric the highest) that would rock; allow me to further separate myself from using Google & support companies who share my values in one interface!
Too true. Corollary of "if you're using something that's free, you're the product" is "if you're using something you paid for, you're still the product". Either way, companies are incentivized to sell your data for profit.
So true. I just had my wife and or teenage daughter watch The S Dilemma on netflix just a few hours ago. Chilling even for a techie like me. More people should be encouraged to watch it.
This is nice and I really appreciate the work of DuckDuckgo for privacy, but unfortunately it is not enough for me to switch from Google Maps. First, you cannot select a custom location as a target, it automatically selects a shop from a search result. I could not select my house or the closest station.
Secondly, there are no public transports, so it is unusable in Tokyo.
That said, both bing and ddg tend to give me similarly disappointing results. They just seem focused on the US and searching in any other country in English, or even dare to use another language the result quality and especially recency drops significantly.
I have used DDG for the past five years and I like it as a hedge to keep other search engines honest but this is not in good taste.
If they want a product blog, rather than using something like this as a name, they can simply do what other respectable companies do and use something like
Are you talking about the footer? A very small percentage of people would scroll all the way to the bottom especially on a relatively long page like this (5583px tall on desktop). Speaking of actually visible branding, there's a DuckDuckGo logo (without wordmark) in the top left corner and that's about it. To most people who aren't familiar with the DuckDuckGo logo, this would indeed look like the site of some privacy advocacy group.
Edit: Didn't notice that the sticky header once you scroll down does say "DuckDuckGo Blog". Not sure why it can't just say that in the actual header; intentionally vague or baffling design?
Call it petty or whatever, this thread is obviously about the entire site, which engages in content marketing a lot (check out all the non-"DuckDuckGo News" posts) without the courtesy of a clear domain name, and not about this particular announcement.
As actuator said in the thread, it should be duckduckgo.com/blog/.
> 1. You only mentioned about it in an edit, it's disingenuous to suggest that I intentionally ignored it.
My comment was talking about the actual header in the first place. And I posted the edit about the sticky header like ten minutes before your reply. But maybe you didn't refresh the page, so whatever. If I was disingenuous I wouldn't have labelled it clearly as an edit, unlike the site which isn't clearly labeled as a corporate blog, until you scroll down.
> 2. It says "DuckDuckGo News" even before you scroll down the page.
"DuckDuckGo News" is only a category among many. A privacy advocacy group blog can't have a category called "DuckDuckGo News"? Hell, many people's personal blogs have Google, Microsoft, Apple, etc. categories.
I pity the fools who rely on these maps for real-time directions. Whatever happened to reading a map and making mental notes? What ever happened to people being able to follow street signs and using your common sense?
How did we become so utterly dependent on this technology that only a generation ago was unnecessary?
Driver-less cars with AI powered navigation coming to your house in the not too distant future. It will plot the most efficient course to get your meat-sack self from your house to the corner store to buy bread & milk....and you don't need to think at all about what your legs are doing. That is if you even still walk. Maybe you are so lazy you will order by the noise-polluting drone delivery service.
People marvel at tech, and don't misunderstand my rant, Google Maps, Openstreet Maps etc are extremely useful in researching places, but, we should not become reliant to the point where we cannot function without them. Otherwise we are becoming mindless zombies in this ever tech-infested world of ours where we are loosing basic skills such as navigating from point A to point B.
We only keep them for debugging purposes and attack prevention. The maps layer (visualization) is done from an external company though: you can change this in the upper right corner to e.g. the original OpenStreetMap or Thunderforest.
Additionally you get nice biking and hiking routes all with elevation, multiple other vehicle types like scooter or truck. And the possibility to export to GPX or using multiple via points. Additionally we try to update the road network as fast as possible (currently max. 1-3 days) from OpenStreetMap data, which is not always the case from other providers. Currently we try to revamp our UI to integrate new features like our alternative routes (you can trigger this via algorithm=alternative_route URL parameter).
Now it is always important to ask the question "how do they make money?" and to understand why we can offer this all for free and even have the routing engine as open source: the simple reason is that we do not have to earn money with the end user. We created the company to push our open source projects (we also have jsprit - a toolkit for solving rich vehicle routing problems) and we try to support OpenStreetMap where we can (e.g. we are the only European corporate gold member of the OSM foundation) and do this via our growing SaaS where we sell subscriptions to many small and medium sized businesses. We do not have VC and so there is no pressure into things we don't want.