I’ve used DuckDuckGo as my search engine for a year | Hacker News Reader

I’ve used DuckDuckGo as my search engine for a year

9to5mac.com

50 pointsdecrypt62 comments

My quick review after more than a year of DDG as my default:

- Technical searches are usually on par with Google, or even better. What made me switch originally was that I got fed up with Google just straight up ignoring half my query, and I've had none of that frustration with DDG.

- I like the dark theme (although I think Google's experimenting with that now)

- DDG is absolutely worthless when it comes to local searches like shops or even just nearby places (I live in Belgium). I always use '!g' for those. I also still use Google Maps.

> - Technical searches are usually on par with Google, or even better. What made me switch originally was that I got fed up with Google just straight up ignoring half my query, and I've had none of that frustration with DDG.

One thing I don't get from DDG is multiple links to the same website, especially stack overflow. I often find information I need to answer my question across several SO questions, but DDG will only ever direct me towards one solution.

Yeah this is why I always end up reverting back to Google. It's just much better for finding answers to coding questions

Another thing for which DuckDuckGo is useless: finding media. It or whatever crawler it uses totally falls for those keyword-flooded sites that have a couple ads and then paragraphs of "wandavision watch free mr robot now free download all seasons season 1 101 dalmatians disney land[...]" autogeneration. I wish they'd do something about all the spam.

I think torrent meta-crawlers might be more what you're looking for. Try the torrentz2 clones

Did you set your location? I also though the results were pretty bad until I told ddg my location.

Damn, that seems to actually work well! Thanks! I'll try it out more over the coming days, looks promising.

It's a good UX quandary too: if there's information that improves the primary user intent (queries) that's privacy-sensitive (e.g. location), what's the most ethical way to permit the user to voluntarily submit that?

It seems to do better with local searches in the US(not surprising) but it's still not very good at that. In terms of maps, DDG maps basically stinks besides the fact that the actual map UI is based on Apple Maps. It otherwise lacks a lot of features that MapQuest even had back in 1998.

When it comes to local searches, i recommend toggling the little switch on the left that limits searches to a specific country/language. I have found that this helps a lot.

That reflects my usage pretty accurately as well. I am rarely on a DDG results page for non-technical topics because I always delegate to other sites via !bang, !g being the most common one.

I still use the !gm bang too; but I'm starting to use the !mq bang more. It's a shame that the !osm bang is broken, because I use osmand instead of google maps.

I’ve used ddg for about 5 years. I find it works about as well as Google. I think Google works worse than it did about 10-15 years ago and that this is just a reflection of the internet itself becoming crappier.

This article doesn't really cover anything that the hacker news audience won't already be familiar with. What is notable, however, is that Duckduckgo has seen a pretty strong surge in usage recently [0], moreso than similar surges in the past.

[0] https://ddg.gg/traffic

I’ve always wondered how the economics of DDG work out, considering they’re basically a proxy for Bing with a bunch of other scraped/API-based data sources they layer on top. Are they paying Bing per query? How can that net out positively?

I suppose at their volume they’ve negotiated great API rates, but their own ads can’t possibly bring in enough revenue to cover the expenses of the API calls…

Gabe Weinberg has previously stated on hacker news that DDG was profitable at about 4M queries a day (it is now over 100M).

The economics are that you charge more for ads than you pay for API calls. That's all there is to it, really.

Some of the links are also affiliated. This is not exactly great for 100% 'unbiased' results but, since their main competition is Google, I think the Rubicon has already been crossed on that score?

> Should Apple buy DuckDuckGo? I’ve seen this comment floated around the tech community for years, and while it’s an excellent idea, I don’t think it’s needed.

As a non-Apple user I can only say: what?

Google pays Apple something like $9 billion a year to be the default search engine. So it seems unlikely.

> Google pays Apple something like $9 billion a year to be the default search engine.

Except in the EU, I guess.

https://www.cnbc.com/2019/08/02/google-to-let-eu-users-choos...

Apple could use DDG as leverage though. If Google stops paying those 9bn, Apple has no reasonable alternative search engine it could resort back to. By quietly continuing development on DuckDuckGo for ~500mn/year, it could be some deterrence towards Google's monopoly that aligns with Apple's own interests.

I think the broader point is that Apple needs a search engine for the same reason it needed a Maps solution: it's basic platform functionality at this point.

And they'd be wise to avoid the Ted Turner mistake: when all your competitors are vertically integrating, don't be the last person left standing without a chair by the time they complete their integration. It leads to an essentially powerless bargaining position.

And like it or not, Google, Facebook, and Amazon are extremely aggressive about integrating their platforms. Microsoft being somewhat of an odder duck.

I use DDG daily but there's one thing that grinds my gears, and undermines their privacy by default statement.

They leak the searches in the URL like this https://duckduckgo.com/?q=this+is+not+private+at+all&ia=web

How is that a leak ? The URL is only sent to ddg server, encrypted using TLS so no eavesdropper can read it. It also stays in your browser history and let you use the back button.

Enable POST requests at https://duckduckgo.com/settings#privacy

That seems like a reasonable trade-off. It's https so it's not visible to people snooping on you, and it being GET means that browser history and sharing links to searches works.

Plus, they do have an option to go POST-only, so if you don't like their already-very-private default you can change it.

Just curious, but what threat model are you thinking of here?

I've always wondered why DDG doesn't go fully open source? Their mobile apps and browser extensions are already open source, why not open source the search engine? Is it just because they don't want to deal with the extra work (they already shutdown duckduckhack), or is there something else that I am missing?

I've tried DDG a few times over the past 3-4 years and reverted back to Google each time until now because, actually, the results are good enough and it's refreshing to see just a page full of search results without all the Google guff.

I switched my phone's default search engine to DDG but still use Google on the desktop. I find the mobile experience for Google search to be much more annoying with the heavy emphasis on AMP.

I gave DuckDuckGo an honest shot for a few months a couple of years ago.

I had to give it up, sadly. The results were just not good enough. I'd love to get away from google, but having less than perfect search had a surprisingly large impact on my day-to-day work.

Just one instance of "I couldn't find anything about x . . . " where a manager responds "I found it in 10 seconds on Google" is probably enough to send just about anyone running back to Google for work-related searches.

I agree that Google is still just a better search engine, and that sucks. But I still try to do my search on DDG first -- if I find what I'm looking for, great! If I haven't found it after a few seconds, I open Google and search there. At most, I've wasted a few seconds. But even if it means Google has some data on me, if I can divert 40% (or more) of my searches to DDG, that's 40% less data Google has on my search history.

I use DuckDuckGo as my default, and have for a couple of years. My workflow is to the run all searches the DDG, then, if I don't see what I'm looking for, I run the same search using the "g!" keyword. If I were more concerned about privacy I'd use the "sp!" keyword. I find that I wind up routing searches to Google a little less than half the time.

What I like most about DDG is that they show more text preview of a site. I find easier to know if a site will be useful or not (google shows very little, and you end clicking).

But as all agree some queries are not really usefull on DDG so I have a bookmarklet to toggle between google<->DDG with the click of a button:

    javascript:void(window.location.href = "https://" + (window.location.host == 'duckduckgo.com' ? "www.google.com/search" : "duckduckgo.com/") + "?" + window.location.href.match('q=[^&]*')[0])

One thing I've noticed I absolutely love is !bang. Being able to quickly jump to code documentation or a sports game is really awesome.

This is the killer feature! I use it all the time for work and life.

Need to find .NET documentation? !dnab

Need to search Mozilla docs? !mdn

Need to find a Debian manpage? !debman

Java docs? !java (or !java7, !java8, etc)

Stack overflow? !so

Google Maps search? !maps

Wikipedia? !w

Amazon? !a

Ebay? !ebay

Last time I checked there were only a few dozens of them.

Phew, now I can search my Earthbound tips right from the browser.

https://duckduckgo.com/bang#bangs-list

!m is enough for Google Maps

!w<ISO> for Wikipedia in a specific language (!wen for English, !wde for German, !wes for Spanish, ...)

!a<ISO> for Amazon in a specific language/country (!ade for Germany, ...)

!py3 for Python docs

!dman is enough for Debian man pages

!wa for Wolfram Alpha

!gt<ISO> to translate into that language, so

!gtes convivial -> alegre

!gten alegre -> happy

I avoid google (even as a backup) because they've replaced displayed URLs with bizarre breadcrumb things. I can't fathom why even a single person at google thought that was a good idea, but no doubt I'm wrong.

I'm sure it was a fun project team to be on though. The "Replace Slashes with Angle Brackets Initiative"

You can use something like https://github.com/palant/searchlinkfix

OK, but... does this mean you don't watch anything on Youtube?

It's not an all-or-nothing thing. For me, I watch YouTube in a dedicated Firefox Container (and everything else in either temporary containers or dedicated), so not much follows me around.

Figure out what you want, and then put together pieces that get you there.

This. Or you can use DDG to search, then copy links you like and use youtube-dl to just download the video.

that sounds good. I need to set up some scripts to do this & also sync into my phone.

Yeah. BTW, on macOS, it would be trivial to hack together an Automator workflow that gives you a Service (ie, a menu item in Safari->Services) that says "Download Video", or "Add Youtube Song to Music", and does what it says.

Basically you'd use the "Get Current Webpage from Safari" block, then in the Terminal

  PATH="/usr/local/bin:${PATH}"
  cd Music/YTDownloads  
  # or wherever you want it
  youtube-dl -q -f 140 "${@%%&*}"  
  # format 140 is 128k .m4a

Then add to existing playlist in music. Ah, if only it were legal!

yup. I just keep a terminal open set to that path. Had just 10Mbit wifi before but that is now 500Mbit. Will have to revisit this solution. Something like a rsync would be nice.

If you want freedom and more privacy, stop using so much proprietary software. How much Apple lover out there talking about privacy or freedom. They are blind.

If you want more privacy, you should use Android rather than Apple??? I think you misunderstand a lot.

Android only comes bundled with Google apps by phone manufacturers. Cyanogen/LineageOS and Replicant are mostly OSS/FS.

IMHO the Apple cult really sucks. Apple isn't really caring about privacy, they are just making a lot of money with false promises.

They are in a position to lose less money from protecting user privacy than other companies, so they’re doing so as a competitive advantage. It’s business, but it happens to work out pretty well for users.

Duck Duck Go uses "more results", instead of proper pagination.

I won't use any website that has such poor judgement in website design.

I am curious here. Why is that an antipattern in said case?

In particular, page based pagination requires a permanent set of elements that are sortable.

Sortable, indeed, you sort based on a scoring mechanism.

Permanent, nopes. If you save page 4 of search page and take a look at it next week, it might be different.

If pagiantion should be properly implemented in a search engine it would be like in elasticsearch where it is transaction and cursor based. But hey, carrying around millions of open transactions is not feasible.

So it seems to be a fair trade off to make an infinity scroll like feature for this.

But again, please enlighten me. I am curious.

Do you ever actually use that? I've come to the conclusion (on all search engines) that if the result I'm looking for isn't on the first page, it's not going to be on the second or third either, so I've stopped even looking there.

I use DDG for all searches, and do occasionally go to the second page of results (and sometimes even the third). When I do, I either find what I was looking for there or figure out what I might want to change or exclude from the search for my next query.

For example, this morning, I was looking up micro fans for embedded electronic devices, and did find some good results on the second and third pages. Not exactly what I was looking for, but close enough that it was worth digging in to the results.

It's not optimal from a sharing perspective, but from merely a searching perspective I really haven't had a problem with it. When I used Google, I would rarely ever go beyond page 2, and I'm sure many people do the same.

You can enable "infinite scroll" for both search results and images. That is not explicit pagination, but it does save a mouse click.

I dunno if it addresses your concerns in particular, but there is a JS-free version with normal pagination:

https://ddg.gg/html https://html.duckduckgo.com

Nope, same crap. Every single "page" has the same URL

I don't see whats proper about navigating to a new page to view more of one list of data. That seems less intuitive although it became the norm.

It does seem like an anti-pattern especially for a search engine.

Do a search for any politically hot topic, in Google, DuckDuckGo, Swisscows, etc.

Compare the results.

Contemplate the dystopian implications of an ideologically correct search engine.

Reply on news.ycombinator.com