Make DuckDuckGo your Chrome default search engine | Hacker News Reader

Make DuckDuckGo your Chrome default search engine

help.duckduckgo.com

305 pointsevolve2k122 comments

OK, but tomorrow when I've forgotten that I've done this and I do a quick search, if the very first thing that comes up isn't what I want, I bet I switch right back.

You get used to typing !g if you want google's results, and then it's automatically run through encrypted.google.com

Also, since we talk about chrome, there are very handful and underused search engine shortcuts built in : simply type the first letters of the search engine in address bar and press <tab>.

I've switched to duckduckgo about a month ago, and indeed I'm regularly making search on google because duckduckgo just can't provide the results I want as well (especially when it comes to non english search). But it's way less a pain that it seems : I just type "g<tab>", and I'm using google. And I still have the high privacy advantage of duckduckgo for most of my searches where the duck is enough.

The bonus of that is that I now frequently use "gi<tab>" to search on github or "y<tab>" to search on youtube.

Hacker news forgets that people need to get shit done, and all of this bullshit marketing is just a distraction.

I found that having the !$var shortcuts to all sorts of sites (including !g for google, !w for wikipedia, and single word strings for many lesser trafficked sites) has made me more productive than a plain google search bar where I often need to add an extra click to the workflow.

As is usually pointed out in these threads, if you're using Chrome anyways, you can do the same thing (and with the same bang syntax, if you want to go to the bother of customizing them that way) with the location bar in Chrome. Start typing the web address, hit tab, enter the query, and it just goes straight to wikipedia, google, etc with no DDG intermediary.

It's slightly different, but you can do almost the same thing with Keywords in the Firefox location bar, and they presumably could support the bang syntax (or whatever) as well.

It's more than "slightly different".

By all means, try and replicate this complete list in your Chrome browser: https://duckduckgo.com/bang.html. I'll wait...

What makes the DDG !bang system awesome is its enormity. You don't look up what !bang to use to search something. You assume it exists, do it, and 99% of the time, you're right. That's a game changer.

I don't write Perl, but I know !cpan exists. I don't even have to look. I don't even know exactly what site it's going to search. But I know if I !cpan twitter, I'm going to be looking at a list of Perl libraries for tweeting.

That's the DDG killer feature. You want to do some kind of site-specific search or lookup, you just type in the most logical sounding !bang and assume it will work. You're simply never going to replicate that inside your browser.

> You're simply never going to replicate that inside your browser.

Why not?

You don't have to replicate the list. Why would you want to? If you've visited a site before and it has a search box on it that matches some heuristic, it gets added to your search engine list. Odds are, far more than 99% of the time, you're going to be searching a site you've already visited before. I agree it is very handy.

His main point has to do with finding new sites about specific topics. A different use case not met by Chrome's implementation.

It's actually a standard, not a heuristic:

http://en.wikipedia.org/wiki/OpenSearch

If the search box is nonstandard, you can also add it manually by right-clicking the search box and clicking "Add as Search Engine".

What makes the DDG !bang system awesome is its enormity.

Please don't tell me it's evil.

I believe you meant "scope", "scale", or "size".

Enormity has a neutral use as well.

> Usage note

> 3. Enormity has been in frequent and continuous use in the sense “immensity” since the 18th century: The enormity of the task was overwhelming. Some hold that enormousness is the correct word in that sense and that enormity can only mean “outrageousness” or “atrociousness”: The enormity of his offenses appalled the public.

>Please don't tell me it's evil.

I am a little confused by your comment - has the word "enormous" come to mean "evil"? If so, maybe this new came about from people referring to the "enormity of a crime". It seems to have followed a similar etymological evolution to the word "gross" as in "gross depravity". But both words just mean "big".

The word "enormity" means generally "evil": The state or quality of exceeding a measure or rule, or of being immoderate, monstrous, or outrageous. (1913 Webster).

It's not a general synonym for "enormous" or "having vast size", except by confusion.

Wow. Maybe it is an American English usage. In British English "enormous" is always used as a synonym for "huge", never for "evil". In fact, a quick google of the word revealed the former usage to be the more prevalent one. For instance, there is a famous children's story called "The enormous turnip" (adapted from a story by Tolstoy).

http://www.boomerangbooks.com.au/Enormous-Turnip/Katie-Dayne...

Obviously, the children's story uses "enormous" to imply that the turnip is very large, not very wicked. Somehow "the dastardly rutabaga" or "the heinous swede" makes a lot less sense in this context.

Edit: Admission - I used DDG (my default search engine), not google.

That's not what he's saying; he's saying that "enormity" isn't a synonym for "enormousness".

I've become very used to doing this. The other handy fact is that if you start typing out a URL you can see the expansion, then hit tab to start the search, rather than having to remember the bang syntax for DDG.

I'll switch between methods though.

I've been doing that for years (Opera was first with this functionality), but simply being setting DDG to the default search versus having to configure all those different keywords, saves me SO much time.

If it was just my desktop, and for some reason it only had one browser, okay. Customization is always cool. But I use too many different computers with too many different browsers in too many different locations that DDG's !bangs makes this really convenient and portable, getting the fast browsing workflow I'm used to in just a few clicks.

It took one search for some css help I needed and I already switched back to Google. I don't have time to see if DDG knows the answer yet. I just need the answer and Google gives it to me.

Strong words coming from someone meta-commenting on a hacker news story instead of "getting shit done."

Lol!

You have to do something on your Pomodoro breaks...

Is really sad that this comment is not grayed out.

That's what I did a few months ago when I tried DDG as default. The learning curve getting used to their design was a hurdle I didn't feel like jumping either.

Just use Google through a VPN you trust. Same threat model, better results.

"^L g<tab> search" and you're searching Google.

I've had DDG as my default search engine. It's been a bit rocky. I've left it in place out of privacy preferences, and used it occasionally, but had actually started getting used to the google search method above.

I'm consciously reverting to using DDG more (it's getting better). Why? Because my privacy and rights do matter to me. I've still got Google literally a keystroke away (or !g from any DDG search, or !sp for StartPage's private proxied Google search).

And DDG's TTY mode is pretty awesome: https://duckduckgo.com/tty/

> And DDG's TTY mode is pretty awesome: https://duckduckgo.com/tty/

How does one find this kind of hidden features? I haven't seen it advertised anywhere except on HN.

HN among others.

I filed a bug about poor integration of DDG with w3m (a console-mode browser) based on default placement of the search button in the tab order. I received a response within the day pointing me at the lite interface: https://duckduckgo.com/lite

... which works perfectly.

That said: these do tend to be somewhat hidden, and I didn't find them through the "Goodies" or "Settings" DDG pages just now. Hrm.

then you type !g after the query (or before) and it sends you right on your way to (https) Google.

Making DDG my default search is the best thing I've done for improving my browser search capabilities. Coming from Opera, I'm already in the habit of using a great many custom search engines, which is nice if you're in Opera, but then Chrome feels very empty. Making DDG the default search gives me just about all of my custom search engines (and more), with pretty intuitive keywords, without having to configure and sync every browser on my system.

ever since ~2-3 years ago, Google is only very rarely my first stop in searching, I so much prefer to go directly where I want.

biggest part of the reason was their second-guessing my query and their results deviating more and more from a simple AND-query over my keywords. it felt like someone replaced my filet knife with a butter knife. (the other part was their instant search result updating and other "enhancements" were super distracting to me).

the only times when I go straight to Google (by which I mean prefix !g to my query), is for a few typical "local/Dutch" or "commercial/consumer" searches, you get a feel for what Google would answer better. Those types of queries are exactly what Google has been optimizing for the past few years, at the great cost of their general and precise web searching quality.

(PS, tip for the Dutch HNers: '!wnl' searches the Dutch Wikipedia. No you don't need it often but when you do, it's nice to have it quickly. I couldn't find it first because I always had it keyed to 'wn' myself, which sends DDG to a weather search site or something)

DDG is nice initiative. The problem is that it does not solve any problem related to privacy because it is based on a matter of trust.

Distributed search, similar to bittorrent, DHT-based designs and the like are notoriously difficult.

I've participated in such efforts, like the Seeks Project [1], Yacy [2], and related initiatives like Unhosted [3], and it takes a certain amount of dedication (and suffering ;) ).

However, I believe it is not entirely impossible that we see a true alternative sometimes. From what experience, what is needed is a slightly better set of distributed algorithms, a business model with the ability to sustain such a technical effort, and a range of features that no search engine can yet offer (because centralized).

[1] http://www.seeks-project.info/ [2] http://www.yacy.net/ [3] https://www.unhosted.org/

Side note: working URL for unhosted is https://unhosted.org/

Which of those is the best for the average joe?

Probably none yet, unfortunately. Seeks is usable because it relies on Google or any other engine in the background, when no results are available among peers.

Yacy is more advanced, uses a DHT, but does not really work for daily use, the accuracy being too low.

Unhosted is distributed architecture on top of the Web really. I've mentioned it because it may be useful and/or an inspiration to whoever interested in these matters.

I've tried to download seeks, but it's linux only. This makes it much harder to build a large user community and to develop a decent business model.

Right, one way is to share one instance with someone you trust (and who runs a linux server), and to use it remotely.

I love DDG and I used it for 6 months but the problem I had is that when I was looking for things related to coding or Linux--it fell short. I would type in an error from Objective-c, Node.js, Java, JavaScript or Scala and nothing helpful would return. This forced me to return to Google which I ended up doing so often that I switched my default search engine back :(.

have you checked out some of the !bang options available? I just took a look and here is a sampling of what you have available.

Sysadmin: !apache !awesome !bsd !chef !codeweavers !cookbook !cve !datasheet !distro !distrowatch !eventid !everymac !exex !exploitdb !filext !fsf !fsfe !funtoo !gdiag !gnomebugs !gnu !howtoforge (!htf) !ip !irp !iso !linode !linuxfr !linuxhcl !linuxmint !lxr !mysql !nginxwiki !openbsd !openbsdmisc !opensuse !oracle !osvdb !parabola !peppermintos !playterm !postgresql !psql !puppet !qth !salix !salixforums !sectube !senderbase !spice !splunk !sqlite !susebug !susepkg !technetlib !timestamp !tldp !trisquel !use !vmkb !vol !wog !zahe

Sysadmin (debian): !debbug (!dbugs) !debian !debianforums !debianfr !debml !debmla !debwiki !dfiles !dman !dpackages (!dpkg !dpts) !dtag !fanch

Sysadmin (FreeBSD): !bsdman !fports !freebsd !freshports !fxr

Sysadmin (Gentoo): !gbugs !gentoo !gentoowiki !gpackages !gwiki

Sysadmin (RedHat): !rbugs

Sysadmin (Solaris): !sunpatch

Sysadmin (Ubuntu): !askubuntu !docubu !docubufr !omgubuntu (!omgu) !ppa !puc !ubottu !ubuntuforums (!ubuntuf) !ubuntuusers !uman !upackages !upkg !uu !uude

https://duckduckgo.com/bang.html

Part of me thinks those are cool, but the rest of my shouts "why should I have to do that?!".

Because Google learns what your preferences are, DDG don't. This is why there is no "search bubble", unfortunately it comes at a price.

That would be reasonable, except that on all my searches that DDG fails on, Google works perfectly fine in private browsing from a different IP.

I didn't know about those. The thing about !bang options is that you have to know something more about the problem to know the appropriate place to look--And that's the whole point of a search engine. The search engine should determine if the best/most popular solution to my problem is on one of those !bang sites or if it's on koders.com.

We're automatically detecting categories on blekko.com now. If you type in a programming question, it should return a category called /programming or /javascript, or whatever the language is you're looking for.

Examples taken from some stackoverflow posts:

http://blekko.com/#?q=Exception%20caught%3A%20java.io.FileNo...

http://blekko.com/#?q=No%20value%20for%20parameter

http://blekko.com/#?q=logger.error('Opps%20we%20got%20an%20e...

We could create a search machine decision plugin for Firefox that forwards tricky queries to Google. Maybe a simple rule like

  .*[\d|\.|\"|\,|\(].*|.{42,}

is already sufficient.

I can support this, but only if it offers real safety as opposed to Google.

So does it? Or is it safe only because NSA hasn't bothered to go to DDG yet, or do whatever they are doing to get Google's search data (cable splitters or whatever)?

They claim not to record any data, so there's no worry about gov't requests. http://donttrack.us/

Oh wow. I didn't know this.

Interesting. This could become a real competitive advantage for them.

No, actually it will not.The vast majority of people using Google do not care whether their data search data is being recorded or not. Heck, most even do not know how Google search engine works from 'behind' let alone worry about the difference between Google and DDG.

So they will really need something more compelling to get the average Joe (who make up the vast majority mentioned earlier), who always get his results from Google's first page, to switch from Google to DDG.

Yet, in the context of search engines, I don't see anything that could be more compelling than always getting the most relevant search results.

You should read their privacy policy: https://duckduckgo.com/privacy#s5

"Also, like anyone else, we will comply with court ordered legal requests. However, in our case, we don't expect any because there is nothing useful to give them since we don't collect any personal information."

Yes, until the NSA requests that DDG allow them to see all searches in real time. You could argue that it says "court ordered" but no one here has any clue what type of pressure the government used to get PRISM through at a major publicly traded company or what they are capable of with a small operation like DDG.

[added] as pointed out below, the NSA could and will just save the searches which would result DDG being just as insecure as google.

Then the NSA only has to save these searches somewhere in place of DDG - we end up with the same result.

What kind of joke is this? You are telling us to use a US-based search engine in a propietary browser made by a US-based company that we already know has given direct access to the NSA? What the fuck.

We don't "know" anything such thing about direct access. We have no evidence other than interpretations of a vague slide that fly in the face other claims.

still... it seems odd to tell people that to ensure their privacy they should change their search engine preference, but keep on running a full-time spying program created by Google.

Do you need yet another individual to become a "traitor" to the US government on this specific assertion before you decide its good enough to "know"?

...and you trust Chrome? Not even Chromium, but... Chrome. Google makes Chrome, you know.

You don't have to trust it. You can sniff 100% of the traffic coming out of your own computer. You can even configure a firewall to block all traffic to Google's netblocks.

All of which is pointless, since you have already installed their binaries.

Even if you have their binaries they'd have to communicate home to transgress your privacy.

edit: Missed the suggestion to block traffic entirely with a firewall - thought the discussion was only on sniffing traffic. Less ridiculous then :)

You're suggesting people that already stated a distrust in Google do a thorough network traffic analysis after every update, so that they know their privacy is violated after the fact? That's a bit ridiculous, really.

It's pretty simple with something like Little Snitch...

"make duckduckgo your default Firefox search engine"

http://help.duckduckgo.com/customer/portal/articles/216441-f...

Yes, I trust that Google has not shipped a binary modification to their open source software to track search terms, which would be (a) illegal, (b) easily detectable enough that someone would notice it eventually, and (c) pointless, since they're already getting the vast majority of users' search terms legitimately. Then again, I work for them, but I would have said the same thing a month ago...

Agreed. I just searched "PRISM" on DDG, and 8 out of top 10 results had some element loaded from a Google server (GA, fonts, jquery, search box, embedded youtube, etc.) Are you using Google's DNS server? Google owns the web. Face it.

How would it be easily detectable?

Pretty lame of DuckDuckGo to use this controversy to gain some cheap exposure.

Besides, there are zero guarantees that they are immune to the kind of accusation currently aimed at Google and co.

It is just linking to the DuckDuckGo support page. You are talking as though they made a blog post saying "PRISM!! Use DuckDuckGo instead of Google" and posted it on HN.

On the other hand, they have been contacting bloggers and media in exactly such a manner. https://plus.google.com/+JeffJarvis/posts/5X7nHcjijsC

If they're not supposed to gain a marketing advantage at the time the problem they're trying to solve is in the spotlight, when exactly is the right time?

I would go even further: how could they prove that they're not an NSA front?

It'd be hard to provide any data to the NSA based on DuckDuckGo's privacy policy: http://donttrack.us/

(Unless, of course, you think they're blatantly lying about what they do and do not track, in which case I'm wasting my energy.)

Why would you use Chrome if you have privacy concerns and want to use DuckDuckGo? Google isn't exactly known for respecting privacy.

Install Firefox instead.

Chrome (or Chromium, if you prefer to have someone else run your builds) is open source. It's subject to inspection. It's got some known privacy issues (e.g., safe-browsing mode sends a file URL and hash which is stored with your IP for two weeks, then your IP is stripped).

But if there's anything really untoward going on and people learn about it, it's going to look extremely bad for Google.

While I wouldn't say that's impossible, it's not a high-likelihood event in my book.

Of course, you're free to use alternative browsers.

Just ... keep in mind that if you're using a WebKit browser you're trusting Apple, and if it's WebCore, you're still trusting Google.

Mozilla's Gecko is still independent, however.

So basically it's got known privacy problems, requires trust of Google and you're winging it.

Lets also add that Chrome!=Chromium. Chrome is based on Chromium. Not all of it is open and it is not compiled in the public eye.

No thanks.

So basically it's got known privacy problems

Most of which can be addressed. There's also a UUID in each build. Privoxy can remove that AFAIU. I don't use Chrome myself, but Chromium.

As I said: you're arguing in the absence of evidence on a pretty widely available and analyzed application. Rather than simply throwing out hypotheticals, a more interesting exercise would be to do some research of posted information to see if there are known issues. Might even make for an interesting HN submission.

I might do that myself.

I switched to DDG about 2 years ago and haven't looked back. My take is they have pretty good clue, are appreciative of security, and they use FreeBSD :)

And Bing.

DDG is a nice search engine for simple research, but I don't use it as much for its search purpose as for its !bang utility.

Being able to search on almost any website by adding a simple keyword is a real plus and I will go even further and say that sometimes it accelerates the search process (say you know you want to search on wikipedia for instance).

However, I have to admit that most of the times I end up searching on google :)

In regards to searching a particular site, is this the same thing as using the tag site:wikipedia.org on Google?

It's better, in my opinion, as it loads the website specified and use the website's search tool directly.

Give it a try, go on DDG and type !w to search on wikipedia.

I've been using DDG as default search engine in Firefox, with Google on fallback under the keyword "g". You can make a bookmark with a URL like

    https://encrypted.google.com/search?hl=en&q=%s

and add arbitrary keywords to the bookmark, like "yt", "g", or "so". Then you can simply enter that keyword followed by the search term in the address bar.

DDG definitely doesn't have the same quality results as Google, but it's close enough in most circumstances for me that the times I have to search twice aren't a huge factor.

I use DuckDuckGoog. Non bash-syntax searches automatically hit google.

The only thing stopping me from making this switch is that doing arithmetic in the url bar stops working... I use that quite a bit. Is this some deep integration with chrome or can it be made to work with ddg as well?

Use AlfredApp for arithmetic ;)

or the command line utility bc

If you use bash, you can also use $(( 3 + 2 )). More often than not, if it gets cumbersome to fit it into Spotlight, I'm going to fire up a Python console.

The fact that Spotlight does arithmetic seems like it's regularly overlooked.

Did google do the math without hitting send? This works for me

https://duckduckgo.com/?q=5*5

Yes: http://i.imgur.com/kfOzmUD.png

I'd rather not have to hit send, and that only works when google search is enabled unfortunately :(

This work only if you have search autosuggest on, which basically send everything you type in the address bar (includeing urls) to google.

Stop using your credit cards while you're at it...

How to make Blekko default search engine in Firefox : http://chepurnoy.org/blog/2012/11/how-to-make-blekko-your-de...

i always make ddg my default... but then ill just use google for a sec and forget to change my search back to ddg and then without noticing ill be using google for the next week.

.... so my project for the weekend is to make a firefox extension that resets your search default every X hours/days.

Have you seen the ! queries on DuckDuckGo? If you add !g to a search, it redirects you to Google. !i goes to Google Image search. I guess there are others.

There are quite a few! And the list is growing... :)

https://duckduckgo.com/bang.html

Like a previous commenter noted, the instant math portion of google is a good feature. Any way to do this with DDG?

Simple queries are done in just the same way: https://duckduckgo.com/?q=42*PI

More complex queries are typically done with Wolfram|Alpha. This goes a long way: https://duckduckgo.com/?q=graph+y+%3D+x%2Bcos^2%28sqrt%2842*...

Why not use Startpage then? It uses Google's results, but without the spying.

ctrl-k "bla" ⏎ is etched in to my brain.

I think the GP meant https://startpage.com/ rather than "the startpage of your browser".

I don't trust Chrome, DDG w/ Firefox is easy and at least they are open about their practices.

Better on Firefox Aurora give it a try

I trust Google than Facebook.

Anyone who tried to find news about the Boston Marathon Bombings on the day they happened via DDG should be aware the engine's inferiorities.

DDG is already my default search engine, and landing page (in FF, mind you).

However, honestly, I prefer Google's results. Has anyone else had this experience?

If you're typing !g more often than not. What keeps you using it?

DDG when searching for reference: yeah; bang syntax is great.

Google for everything else.

!g everything else

How does this help? Isn't DDG still located in the US?

Chrome passwords have an export feature, right?

Better Privacy VS Tailored Search.

Using DuckDuckGo in Chrome is like asking for privacy while walking around naked in Times Square. Chrome has 100% unrestricted access to anything your user can access on your PC, no matter what search engine you choose.

Here's a full list of what data is sent back to google: http://lifehacker.com/5763452/what-data-does-chrome-send-to-...

Do the autocomplete results still go to google? If so, then enabling DDG on chrome is indeed fruitless.

Just as an FYI, new versions of chrome also sends back the full URL and the hash of any file you download or save. I bumped across this when I was trying to download light table and it din't let me open it. After a bit of research I found file scanning to be the new addition to chrome's malware checks. With all the NSA debacle going on now, I am feeling even more paranoid. I love chrome but I will be switching to Firefox or IE10 soon.

Chrome is the most invasive piece of spyware you could ever install.

Unless you turn off some 10 or more options, it literally phones back your entire online activity, in or out of incognito mode, in or out of any google service.

No, it doesn't. I've detailed how safe browsing works in Chrome, Firefox and Safari works before[1]. tl;dr only partial hashes are exchanged, so you can't reverse it even if you really wanted to.

[1] https://news.ycombinator.com/item?id=5588362

Yes it does. Check this page out [1] and this stackoverflow question.[2]

"If a file isn’t from a known source, Chrome sends the URL and IP of the host and other meta data, such as the file’s hash and binary size, to Google."

[1] http://blog.chromium.org/2012/01/all-about-safe-browsing.htm...

[2] http://superuser.com/questions/387724/how-to-disable-downloa...

I cannot say enough good things about Firefox Aurora. I've been using it as my primary browser for a few months now, it's awesome.

I've always used Chromium, but I'm also planning on switching to Firefox soon.

Also, IE10? You don't trust Google but you trust Microsoft?

It's only really meaningful if their operations and management are outside the u.s. Is this the case? Are there search companies with their engines in Switzerland or even the EU?

Reply on news.ycombinator.com