Google started this really annoying habit of including results that don’t have the words you are searching. You can get around this by including the word in quotes but it’s frustrating to spend a few minutes looking items returned by Google to not find what you’re looking for to realize they showed it to you even though it didn’t have your search terms.
I bet this is great for current events like “when is Star Wars 1 playing” when the person probably meant Star Wars 8. But sucks when researching items or searching for a paper or particular product.
I started using Duck Duck Go and it works great. The privacy is an added bonus.
My biggest search wish is an option to eliminate any results from online shopping sites. Hell, I want a "give me the old web" toggle that eliminates shopping sites, social media, and any corporate websites so I can just find the hand-written HTML from the weirdos who'll tell me how to change the clutch on a 1969 MG Midget.
All search interfaces should support "Don't show me this again". All of them.
Imagine having a roommate who doesn't remember you want to avoid sugar and are allergic to nuts, yet keeps making PBJ sandwiches when it's his turn to make dinner. That and you're getting effin PBJ for dinner.
Web search: Don't show me anything from this domain again.
Google: You are so keen to remember things about me that I don't want you to remember. But you don't bother to remember what I reject? Of course, because I am the product, not the customer. This feature might run afoul of your relationship with advertisers, yeah?
Yelp search: You've suggested this place before, I spent the time to read the details and reviews and rejected it, but you keep showing it to me on future searches, and because as a human I can't remember all the stuff you've showed me before, I waste time considering it all over again. You even show me places I rated 1 star. STOP!
Maybe DuckDuckGo could add this feature? Please? Pretty Please?
I find, when searching for info on a relatively uncommon bug DDG will give me a page that starts with a relevant stackoverflow answer then continues on to list a bunch of sites that scrape SO and repost the content. It's extremely annoying.
This is my frustration as well. Today's web (as seen through search engines) is just content marketing.
And to reach such a page, you have to go through ads on the results page, then a newsletter popup (sometimes two, these days), then ads all around the content (which is still marketing, remember).
I'd like to see a search engine where a trusted/verified set of users are allowed to categorize websites and the categories thus derived can be used by everyone to help filter their searches.
So when a categorizing user does a search and clicks on a result that takes them to best-auto-parts.co.uk, they click "Shopping" on their categorization toolbar. After one or two matching categorizations, that site gets excluded from searches which have "shopping sites" turned off.
You could also turn the categorizing thing around and decide that you only want to search within web forums (phpBB systems and similar), where you'll often find people posting rather esoteric knowledge about whatever esoteric car you're searching for.
You are describing Yahoo, ca. 1995. The problem is that new users have no idea how to navigate the categorization system. (Remember using the dewey decimal system? It had the same problem, except that your local librarian was there to help.)
Yeah, I remember the old categorization thing. I'm thinking of a much more high-level categorization: Here's sites selling shit. Here's web forums. Social media. Blogs. Commercial sites.
Honestly what I'd really like is a toggle that I can flip over to say "only show me sites that aren't trying to sell me shit", and then maybe I can read an honest review about the tent I'm interested in without 50 bullshit blog entries that just copy-pasted the manufacturer's copy and added an Amazon affiliate link.
DMOZ[0] was that catalog, more or less, and was used as you describe, more or less. It finally died last year, although I think it was quite dormant for the last 10 or so.
Basically you want Yahoo but with a Wikipedia-like hierarchical tagging. You need a reputation system though so the editing isn’t just seo’d getting us to the same situation we’re currently in.
If Facebook had an api, you could filter based on the tags just provided by your contacts or something like that.
I'd use a search engine that removes any result that includes an affiliate link. Some shopping sectors are so crowded with affiliate links that it's impossible to figure out what is real or not
I use DDG all the time now for personal use, but I still use Google at work.
Try this one weird trick to get only what you searched for on Google:
Click on the "Tools" link (under the search box) and change "All results" to "Verbatim" and I think (I dont know for sure) that this will only search for exactly what you typed.
I experience the "overgeneralized results" issue multiple times every week for maybe the past year. It frustrates me every time. If I have to click Tools > All Results > Verbatim every time, frankly I am ready to switch search engines. I am afraid to say that Google has really let itself go.
Is search still the core of their business? Or does most of their income come from things like advertising, AppEngine, google apps for business and android-related stuff? I too feel that google search has been decaying, and saw many people here complainin about the same thing. Maybe they don't want to bother that much with it anymore if their focus has moved to stuff that they actually sell.
Research shows that you can show people intentionally screwed up results, brand them as “google results”, and they’ll prefer them to real google results with swapped branding.
The amount you have to screw up the results to get people to prefer the real results is greater than the difference in performance of the major engines.
Conclusion? Search quality is a cost center at google at this point in the game — they have no economic incentive to improve it further.
Is it though? They have sooo many sources of input, embedded ads everywhere, social buttons, gmail, calendar, documents, photos, android, blogger, youtube, google+, and probably others that IDK, that search is just one---and probably a less important one---of those. As a means to show ads, again, they have uncountable spots all around the internet that they can place ads, and many google apps that one might be using, that the search results page is again just a slice of a big pie.
Or maybe they figured that most people do very simple searches nearly all the time.
I know they earn a lot more from ads on google properties than on third party sites, you can see it in their financial docs. It's something like 80% of their revenue.
So we know embedded ads all around the internet aren't challenging it in any way.
That said I don't know what percentage of their "google property" ad revenue is youtube, gmail, etc. Maybe they're cash cows?
I'm inclined to think it's still mostly search though simply because it's the most valuable source of lead. One where the consumer has buying intent.
Searching for a new car Vs trying to watch a show and having a car shown to you before it.
I can't remember the last time I used a general purpose search engine with an intent to buy. The closest would be an intent to comparison shop, but then I'm explicitly searching for feature comparisons, not stores to buy from.
Yes. But then you can't filter by time range. In their infinite wisdom, Google has made it either-or. Either I can filter by time, or by verbatim. It drives me nuts!
If you think you were frustrated before, just wait: you get to issue about three searches before you have to solve captchas to see the results. Google thinks only bots want to use those features simultaneously.
Seriously? This would be a deal breaker for me. It would also be nice to be able to save preferences so that I only ever get results from the past year unless I specify otherwise. It just takes additional clicks and is a big pain from the search engine that's supposed to be user friendly.
Google is a sitting duck for innovation, why does it not suggest scopes for search (paris = person / place), or show me a graph of results over time, or let me exclude pages with X% similarity and show me the most likely original source, there is so much room for improvement, exciting times...
Seconded... I use DDG or Google + Verbatim mode. I set DDG is my browser default, and use `!g` to hit google if I don't like the results. (or !gis for google image search)
Thanks. This is really helpful. Comically I googled and DDG’d for how to do this to no avail. And last time I checked under tools I didn’t notice verbatim as an option. But tools is different now. The shifting UI should make me recheck.
Even though I can’t set this for a default, I can at least choose it when necessary and it’s faster than quote escaping terms.
I recently tried DDG again after reading the quality was approaching Google. I was surprised to find the results are actually much better, for exactly this reason. The results I get from Google are often filled with answers to distinctly different, but much more popular searches; sometimes even after adding quotes around terms.
Another nice feature of DDG is that the '-' operator works very well.
I hate this so fucking much. Seems like every other search 70% of the top results have one of my necessary keywords removed. And I don't want to switch to verbatim because I do appreciate the fact that they search synonyms and don't necessarily want to remove that.
This is why I can’t switch back from ddg to google. (Not that I really want to —- the only google feature I miss is the calculator, and I can get that with g! searches.)
Ddg has been slowly cranking up synonym search, but it’s much better executed (triggers less often) than google’s implementation.
Ironically, the appeal of Google over competitors when they were just starting out was precisely that you could find what you were searching for without having to add a bunch of symbols to tweak your query.
This makes me think. Why not add a Google Science or Google Dev with those features removed. Basically Google Scholar but also showing stackoverflow, blogs and GitHub code/gists?
Also Google without country redirection, my pet peeve. I can't use Google in private browsing mode any more since it insists redirecting to German results (yes, I'm currently physically in Germany), so I've started using DDG too.
I used this as default before I switched to DuckDuckGo completely. For a software developer google.de results are pretty useless. Switching the language to English helps a bit, but ncr results are a lot better for software and tech topics.
Also, if you want to break out of your own bubble and get non-personalized results, Google is pretty useless, because of the annoying data protection pop up.
Pretty much the only times I use Goggle is for local search , otherwise I’m pretty happy with ddg.
I believe ncr doesn't really work for this purpose anymore as searching on google.com will give you results for the country you're in without redirecting you to a different TLD.
Preferences works in regular mode, but I've noticed I mostly browse in private mode these days. Opening a private browser window has just become a habit I suppose, and probably not a bad habit either.
I do use NCR to switch to regular results in normal browsing mode where the selection sticks, but I don't know how to do that for the Firefox search box in private browsing.
In my experience Google is usually giving more relevant results to software development related searches, so I would like to retain that at least as an option.
A long time ago they changed it so that using '+' would cause it to search within "google plus" entries. They did this completely silently and the sad thing is it took people quite a long time to figure out that the meaning had been changed.
Yes, DDG does it too. The other significant difference I've noticed, in my admittedly short-but-intense periods of DDG use, is that unlike Google, it won't block you with a CAPTCHA for "searching too hard"; but sometimes it just runs out of results.
Looking through too many search pages, or using queries containing quotes and other "advanced" things like site: or intext: or intitle: operators tends to get you blocked by Google pretty quickly, which is extremely infuriating --- and the point at which I switch to DDG or something else for a while.
They started doing that recently :-/ I guess the appeal to ape leader is strong, even if leader lost its ways. But I am running out of options for search so have to live with it...
I think there's actually a better example of the usefulness of Google's broad search embedded in there: Somebody writing "Star Wars 1" most likely intends Episode 4 rather than Episode 1.
I agree, this is extremely annoying, Google should add an option - for their users who know what they are doing - to always do a verbatim search.
I also still detest their mobile first format on iPad, again why on earth don’t they let their users choose Desktop format, if they prefer this. I have to use slower than Safari iCab Browser to get around this.
The minus option did once filter out results with that word, and could be amazingly helpful to find less used meanings of primary search terms. But it did go away about the same time as the + operator.
Depends on your browser, but for example, in Chrome, you can set your default search in chrome://settings/searchEngines for the omnibox search. Mine is:
"Google started this really annoying habit of including results that don't have the words you are searching."
Queried a one word string that is probably in hundreds of pages.
Got 5 bogus results.
On the plus side, I noticed ddg has apparently fixed an issue they used to have with prefixing result urls with a ddg proxy url, as other search engines are known to do.
To disable this, previously the user had to turn on Javascript. I never use Javascript, and I got the urls unprefixed, as they should be.
search: test2 (no quotes)
23 results
15 had test2 inurl
3 had test2 intitle
0 had test2 inpage
5 were not found inurl, intitle or elsewhere
see below
1. http://www.test.com/
2. https://dndprojects.anheuser-busch.com/mobile
3. https://www.icardnet.uillinois.edu/public/
4. https://www.dmv.ca.gov/portal/dmv/?1dmy&urile=wcm:path:/dmv_content_en/dmv/pubs/interactive/tdrive/clc3written?lang=en
5. https://test.salesforce.com/
cname is "test.l2.salesforce.com"
Some thoughts from a search engine author:
"For most popular subjects, a simple text matching search that is restricted to web page titles performs admirably when PageRank prioritizes the results (demo available at google.stanford.edu). For the type of full text searches in the main Google system, PageRank also helps a great deal.
"Our compact encoding uses two bytes for every hit. There are two types of hits: fancy hits and plain hits. Fancy hits include hits occurring in a URL, title, anchor text, or meta tag. Plain hits include everything else."
"We chose a compromise between these options, keeping two sets of inverted barrels -- one set for hit lists which include title or anchor hits and another set for all hit lists. This way, we check the first set of barrels first and if there are not enough matches within those barrels we check the larger ones."
"Google considers each hit to be one of several different types (title, anchor, URL, plain text large font, plain text small font, ...), each of which has its own type-weight."
Years ago, but years after they launched, I heard they hired a large number (100s? 1000s?) of editors to manually tweak page ranks. Basically, page rank is easy to game and doesn’t really work anymore.
This is great! I have been waiting for an extension such as this which: (1) is seamless, (2) doesn't break the web, (3) doesn't require too much tweaking, and (4) is offered by a company I trust. Mainly in order to install on friends' and family's devices who aren't tech savvy – this appears to be exactly that. Thank you. Love the UX (sans below), nice touch integrating with ToS;DR.
Feedback: The "enhanced from X to Y" is not perfectly clear to me. I'm guessing you're saying "we were able to upgrade the rating to Y from X". Correct? The badge seems to be showing the upgraded rating rather than the "actual" rating – this is not what I expected. I expected to see the privacy grade of said website, not the grade received by blocking scripts and/or redirecting to https.
Question: Will the extension for Safari be available from the Safari Extensions Gallery[1]? This would be even better, avoiding (said friends and family) having to "Trust" the safariextz and get the disclaimer that it isn't from the gallery. (I currently only see your old search extension in the gallery, will they, perhaps, be merged?)
Use DuckDuckGo every day, but wish it had a date range of one year as a search filter like google, needed for programming searches where the landscape changes rapidly...
This has been a long-term struggle to get right, though I really see the light at the end of the tunnel and think it may come sometime this year, hopefully sooner than later.
For others who may not know though, we do have last month -- drop down above results.
I just wanted to take a moment and write my appreciation for your fantastic work. Services such as duckduckgo are needed now more than ever. The fact that it is stable and responsive enough for me to depend on it 100% every work day is amazing
Curious, why is that of all things so hard to implement compared to the gargantuan task of searching the entire internet? I know nothing about the field of search or big data but it seems to an outsider like it would be almost trivial compared to the rest of what you do.
However, there can be situations where it gives you malicious results instead.
Some months back a client of mine was searching for some insurance forms and included the year. The actual forms don't change often and the insurance company website did not include a year, so what she got instead was a bunch of malicious results for pages that included all of her terms plus a year. She was using either Bing (because it was the search engine on her browser on Windows....) or Yahoo, but a very similar set of results came up on DuckDuckGo when I checked there. Google's results were clean.
I did report the issue to both Bing and DDG, but I haven't gone back and tested whether results have improved.
I applaud the initiative, but how does the Firefox extension differ from Privacy Badger or Disconnect?
Does the extension learn trackers as it goes, like Privacy Badger? What does it offer over something like uBlock Origin with the appropriate tracking blocklists?
Also in Firefox, I now have two "Search DuckDuckGo" entries in the right click menu, one from being the default search engine, and one (with an icon) from the extension. They do the exact same thing, so why have them both?
We're trying to put all the privacy essentials we can make seamless -- tracker blocking, upgraded encryption, private search and more to come -- all in one package, across all major browsers and platforms. In this respect, on any major mobile device or desktop browser, you should be able to look up DuckDuckGo and with one download get seamless privacy protection as you search and browse the web.
With regards to other extensions, we found that they generally lack some combination of all the essentials (e.g. missing encryption, private search), aren't totally seamless (i.e. break some of the web), or aren't available across all major browsers and platforms.
With regards to tracker blocking in particular, we would like to be as comprehensive as possible while not breaking the Web, and are close to that with this initial launch, utilizing some open source lists including Disconnect and Easylist. We are not currently using the machine learning aspects of Privacy Badger, however.
There are more nuanced UX differences, however. In the UI we're trying to move away from 'x trackers blocked' and instead group trackers from networks together, trying to identify the umbrella company and purposes.
Second, we're grading each site based on its privacy measures (including privacy policies with help from TOSDR), and telling you at a glance how protected you are on an A-F scale, based on what we could do (e.g. block trackers and upgrade encryption).
Thank you for the feedback on the right-click menu. We will look into that.
What I gather is that the new DDG extension functions roughly as a replacement for Disconnect/Ghostery (or specific lists in uBlock Origin), HTTPS Everywhere and ToS;DR, as well as adding DDG search to the browser (if applicable)?
As a one-stop extension, that's a pretty good deal. In addition to the learning aspect like in Privacy Badger, have you also considered canvas blocking/scrambling, link referrer cleaning and possibly a function similar to Decentraleyes, to prevent big CDN tracking?
How are you dealing with the "temporary disabling" scenario?
I have Disconnect and uBlock, but then it's a pain to use sites like Quidco, that rely on tracking. Ideally one should be able to easily say "disable all protections for 20 minutes" or something like that, do what needs to be done, then after re-enabling protection the slate (cookies etc) is wiped clean (or ideally brought back where it was before the suspension).
The way I end up doing it, I must keep a browser (typically Safari or Explorer) fairly unprotected just to deal with those transactions, and use the hardened-up Firefox for everything else. That's suboptimal.
No issue here on 58, with uBlock Origin and Privacy Badger. Maybe HTTPS Everywhere conflicts, as it provides the same functionality as DDG (forcing pages to use HTTPS).
One privacy essential is a built-in vpn/proxy like Opera has done with its browser, this makes it slightly harder for some fingerprinting ad trackers to follow a user around.
Proxies can be useful, but I'm not comfortable with sending all my traffic over a centralized proxy operated by a single entity. Just like Cloudflare, the entity that runs the proxy gets a free pass to MITM everyone!
A decentralized solution that can be easily configured to send traffic over any third-party VPN service, HTTP proxy, or ssh tunnel would be much cleaner. Bonus points if the proxy seamlessly kicks in only when visiting a site with a low privacy score or when browsing on public Wi-Fi.
slightly off topic but will the extension work for mobile as well? [off topic bit:] because I've noticed the mobile ddg experience isn't as great as on desktop: the !bangs are harder to type, image search on mobile is frustrating, etc... (and even more off-topic: have you considered partnering with tin-eye for reverse image lookup?)
On mobile, we produced apps that have all the extension functionality, plus additional browser functionality, and more privacy browser features like a one-tap erase data button (we're calling the fire button).
Thank you for the feedback on image search -- tineye is something we've considered, though haven't implemented yet.
I'm not really interested in a different browser, though. I like Firefox, and I like the consistent sync to my other devices running Firefox.
Right now the DDG plugin doesn't work on Firefox for Android, it just says "we don't filter special pages" or somesuch, probably because it tries to filter its own tab? (Firefox Android pops up extensions in separate tabs)
btw I just tried image search again after being scared off a while ago because it was buggy, and it's in a much better state than it was back then!
Love how you guys show you don't have to sell out to be able to thrive, and that caring for freedom shouldn't come at the cost of design or functionality (quite the opposite!)
Disconnect" was killing/blocking the Google's captcha so it had to "go". I am using AdBlock+ (a slightly earlier version when Ctrl+Shift+V still works), NoScript and a BIG hosts file from someonewhocares.
I am more than happy to see and adopt any efforts that would reduce tracking and props to DDG for doing the effort!
I will of course be checking/testing their app on iOS as well!
Let me use this thread to lazily look for simple advice.
For privacy, I'm now using: DuckDuckGo Plus, First Party Isolation, Smart Referer, HTTPS Everywhere, uBlock Origin, Privacy Badger, Canvas Defender and Decentraleyes. There's also NoScript installed, but disabled (waiting for its clickjacking protection to return in WebExt version, although I might switch back to whitelisting JS sometime). Plus there's also Firefox's Tracking Protection enabled all the time.
Anything worth adding? Anything that can be removed for sure, cause it just duplicates something else from this list? Anything should be replaced? (if yes, then why?)
ClearClick is what I'm interested in, not script blocking, at least right now. WebExt version doesn't have it now, but I'm keeping it installed, so it just upgrades when it's added back.
Firefox bypasses right-click blocking when you press shift.
What it doesn't bypass is websites placing transparent things on top of what you wish to right-click on (for example images on Twitter). For that I use Stylus to add CSS and move things away.
It's not about right-click bypassing, it's exactly about transparent things. ClearClick checks whether the image of what you click matches the image of element that's being clicked, shows you the preview and asks if you really want to click it. The best pre-WebExt feature of NoScript.
Are you using Firefox Quantum? I used to whitelist cookies on a per-site basis, but WebExtensions no longer appear to have the necessary access for that, and the built-in settings page is too much of a hassle. Would be interesting if uMatrix has a way to provide the functionality anyway.
> as an additional benefit it helps fight the ad industry actively
What does that mean, and why would I want to fight the ad industry actively? (I'm being serious, I don't like the tracking of the ad industry generally resorts to online, but I think blindly assuming the whole industry has no value whatsoever is very shortsighted)
Edit: Man, I hate when I write 300+ words on a a reply only for the parent comment to be deleted prior to my getting to post it. :/
> For privacy, I'm now using: DuckDuckGo Plus, First Party Isolation, Smart Referer, HTTPS Everywhere, uBlock Origin, Privacy Badger, Canvas Defender and Decentraleyes.
If you're interested in this, I suggest the new web browser Brave. It does a great job of cleaning up all the ads & protecting your privacy. I also haven't ran into sites that block me yet like I do when using an ad blocker on Chrome. You can choose the search engine you want to use unlike this extension. It's also built on top of the same code as Chrome so you get all those benefits as well. Brendan Eich (JavaScript & Mozilla) is the CEO.
Slightly off topic, Brave also has an interesting feature as well to help support content creators. Since a lot of content creators rely on ads, you have the ability to add a small amount of funds to a crypto wallet that Brave can use which get distributed to the websites you visit the most over the month. You can enable or disable this for specific websites as well. I think this is a great idea but it may have a hard time catching on. I think people will have a hard time adding money to a virtual wallet to pay for something they "think" they are already getting for free. At the same time I love the concept as it is a way for you to reward people for good content & create a disincentive for bad content.
Brave also started to get annoying by always asking for donations to content creators. So I think i will drop it and continue FF Quantum + uBlock + noScript
How about a simply browser extension that let's you donate to page / site X? This would also be great on GitHub and GitLab.
You could buy credits in bulk so the credit card transaction fee only happens once. And then, if the page has an account, dole out those credits as you surf.
Probably not massive amounts of money but better than the ads based model. At least there would be a option.
Both are close, but not what I'm looking for. I want...a browser extension...that recognizes (via code on that page) a site/page that's registered within the network...the extension icon turns green to let me know...I click the icon...the dropdown shows me my current number of credits...and let's me donate to the current site...
The site doesn't pay anything to be in the network (sans maybe a low sign up fee to mitigate spam). The members pay, tho' I'd have to think about that model.
What I want is semi similar to an old school PayPal button but more micropay friendly (low cost) and is more (social) network-centric.
No can do?
p.s. FWIW this is something I imagine say GitHub doing for starters and then branching out from there.
DDG as a search engine is just good. Its been my default for a long time. I wish Mozilla would make it the default so more users could be introduced to it. Why hasn't this happened yet?
Ads of course. Bot not as greedy as google. Not tracking every fingerprint of you. They just show sponsored search results according to your search terms at that moment, in contrast to your whole browsing history/habits.
I use ddg as my main search engine but disabled ads because I can't stand them. I wish they'd let me pay for the service. "If you're not the customer you're the product". I want to be the customer.
Advertising. But not via tracking and personalization. They show ads to you based on the keywords you used to search and don’t follow you around the web.
I tend to agree, but at least a few times a day have to resort to using !g to actually find a decent result. It's usually Angular/Symfony related but just last week I couldn't find the website of my local tennis club at all, not even by exact name. On google it was the first result...
The answer I suppose is almost always found on stackoverflow.com anyway, so why not go to SO search directly? Because SO search isn't up to it of course.
Why don't they improve their search system? I think they're really leaving money on the table, or maybe it's a job for a third-party SO search app for a fellow HNer (not that there aren't already enough SO scrapers showing up in searches).
For me DDG results are vastly inferior to Google's. If there was a search bar addon that defaults to DDG with the !g bang I would happily set it as default though
Convenient redirection? If you do a search on DDG and don't get a viable result, move to the beginning of the search field and add !g and try again.
Depending on their logging, this may also allow DDG to see what kinds of searches are being done and then redone/redirected to Google. In aggregate, that may provide useful information on areas where their search results aren't good enough.
Just to clear things up: it really doesn't matter where you include the bang. The beginning, the middle, or the end of the search query, the result is the same.
When looking at the search results H, left or right arrow, and type in the bang.
Use Firefox and Container Tabs. I have set up a Container for Gmail and use tabs for Google Calendar, sites that use Google for authorization, etc.
My default window (and other Container Tabs) doesn't have a Google account attached to it. So I see "Sign In" on the top right when doing Google searches.
you can use !s instead, which goes to startpage, a proxy for google search results (i don't think it will completely hide you from google, but it's better than a naked !g search).
It's really odd - sometimes it's great or even better than Google, other times it just does really weird stuff. Putting porn at the top of the results, totally ignoring one aspect of a topic and giving 2 pages for only one meaning of the word (I assume Google has special handling for that?), when maps are shown for geographic queries seems to follow no pattern at all, ...
I'm using DuckDuckGo every day, but I wish DuckDuckHack was easier to contribute to. Maybe something like letting users load their own modules for local use, and public repositories? I'd like to help improve instant answers and widgets (like json pretty print).
It wasn't really difficult to contribute to them about a year or two ago. I've contributed a couple of cheatsheets (for example, clicking on the info icon when searching for "nmap cheatsheet" will reveal my GitHub username), which were, at the time, considered a low-hanging fruit way of contributing instant answers.
Later on, I was later on added to the DuckDuckHack Contributors team[0], which was border-line spamming, up to the point where I had to redirect email notifications from the DuckDuckGo organization to my non-primary email address. Occasional email is fine, a dozen or two dozens of emails in my inbox almost every single day was not.
And finally, I am aware that there was some grand change about how people contribute to DuckDuckGo introduced relatively recently, which makes my experience a bit outdated.
The combination of their actions after I've contributed to them made me pretty uninterested in the idea of contributing more instant answers. Nowadays, I only suggest bangs from time to time, which is as easy as filling out a form[1].
Sorry about the experience you had with the org. DuckDuckHack is indeed in maintenance mode right now as we couldn't figure out the best contribution model that scaled with the search engine, though are still thinking about how to do that.
All of these new apps and extensions are open source though, and we'd welcome any contributions.
Are people mostly positive about this? I was already using DDG - then I get an unexpected pop up asking for permission to rewrite everything. I think we should develop a reflex to routinely refuse to give popups permission - so now I've lost DDG in the search bar.
I would much rather that it stayed as a non-evil search provider rather than bundling ABP/Ghostery. Maybe I'm in a minority wanting separation of concerns?
Or maybe it really is better than ABP and Ghostery and I've missed it by being put off by the delivery mechanism?
I'm sorry for putting you off. The intention is to have all the privacy essentials in one package, and make those essentials each best in class. Over the coming months we will explain more about how we're doing so and why we think that is the case. Right now though know that it is doing more than just tracker blocking and private search: it is also upgrading encryption where possible and exposing poor privacy practices when known.
In any case though, I understand if you feel you are covered with other extensions and just want search. In that case, you don't really need a DuckDuckGo extension and can set the search via these instructions: https://duckduckgo.com/install
I'd still be nice if the omnibox search can be changed. That's the reason I've removed the extension after 15 minutes. I have my own search service with a couple of keywords I need all the time. The fallback/default search is ddg anyway.
Edit - Chrome only issue for which the devs don't have good options.
I was very disappointed to discover that the app forces you to use DDG for search. In Chrome there doesn't appear to be a way to use Google as your default search engine while the extension is active. I like its other features but Google search IMO is much much better than DDG and don't want to be forced to switch.
This is unfortunately the way Chrome works right now. If you revert search settings they disable the whole extension. We are actively trying to find a way around this behavior.
I don't think so in Chrome, no, though we are actively trying to find a way short of forking the extension, and if anyone has any ideas, please feel free to reach out to me.
However, we do suggest generally to switch browsers if privacy is a primary concern. We recommend and use ourselves a number of good browsers: Firefox, Safari, Brave, and Vivaldi among them.
Immediately uninstalled because of exactly this. The "We rate websites on privacy" value proposition didn't quite mention "(and we will force you to use our search without asking)".
That does not justify this in my view. "It's easy for users to change their search habits to get the same experience they had before, so let's just change the default search for them".
Just to add to the other comments already addressing the !g option: I use DuckDuckGo exclusively but I almost never use it's actual search results as a Google replacement for precisely this reason. It's just that I use !w, !imdb, !m, !n, !ups & friends so often, than it's worth it to just prefix my search with !g when I do want traditional web search. But now I've stopped using Google as a shortcut to get to what I already know I'm looking for but don't want to type the full address. I really just use DuckDuckGo results when I do want the privacy especially.
Improving privacy is great, but I find the extreme approach to erase all history taken by DuckDuckGo and others make them difficult to adopt.
I often use the history feature to re-find sites I remember visiting before. Bookmarks don't solve this problem because even if I tried to remember to bookmark anything that might be important later, I wouldn't be able to nor would I always be able to make that judgement accurately.
So, what I like to have instead of the nuclear option (erase all history) is a browser that protects my identity but also my history (e.g. behind a locked gate) that I can use always without having to think about so I don't have to choose between privacy or history.
Some sites have 100s of fallback ads and trackers (usually just domains and subdomains), once tracker 1 failed to load, it tries to load 2 which is the same file on a server, but with different URL. They have literally 100s or 1000s of such fallback URLs.
I'm just grateful that they still serve ads over separate domains, instead of rigging up the ad tech on the server side and embedding tracking into the "functional" Javascript.
I get more worried that once Wasm hits, the only way to tell if you're being tracked will be to monitor individual HTTP requests, which even then might become impossible if everything is sent to and from opaque CDN domains.
I imagine the only thing currently standing between us and this nightmare is that people don't want to actually work with any tracking on the backend; easier to embed it in the front-end and let the vendor put whatever they want in the script and serve all the requests.
Only a matter of time until Cloudflare starts offering ads. They can embed them at the edge with no/minimal JS and links/image files of the host domain.
This is why the internet won't get better until ads are cut out like a cancer.
It is more profitable for them to keep adding layers of ads and trackers than it is to actually build something useful that people would be willing to pay for.
I was curious about the "fake" site they used in their example, creepysite.com. I was pretty surprised to find that it redirects to google.com by an HTTP 302 redirect.
I wonder how DDG's new privacy tools compare to uBlock and Brave.
The Terms of Service; Didn't Read [0] project is super cool. Looks like their information is a bit dated though. Hopefully teaming with DDG will give them a boost.
Ukraine region not available any more: while "ua-uk" URL parameter is still mentioned on search params page[1], it's not working. I didn't get the answers on DDG forum (now closed in favor of reddit), neither on /r/duckduckgo. I suspect this could be connected somehow to Yandex block. Anyway, currently can't use DDG for local search, sad :(
Thank you so much! We really appreciate the support and are glad that you're enjoying the search experience. This announcement should give you more privacy protection when you leave our search results pages.
Privacy is cool and all, but I switched because Google was very obviously suppressing results. I don't want to be in a political bubble, completely unaware of events that might not be pleasing to Google employees. I'm not Google's small child.
I notice AdNauseam blocks your adds even when I have it set to allow non-tracking ads. What sort precautions does DDG do to protect users who use your ads?
Our entire search results page is anonymous at page load, and our ads are non-tracking in the sense that they are just based on that one search page request and have nothing to do with any other personal information (as we don't have any). It's unfortunate that adblockers are blocking our ads, but I understand that our more privacy-aware ads are extremely rare.
One thing I often add to my search terms is the site: tag to restrict results to, say stackexchange, reddit, etc... In the past it's been a necessity to get non-spam results on DuckDuckGo. I'm not sure how needed it still is.
I'm concerned that I'd be handing over data about my browsing habits to DDG and its partners in order to use some of the features described in the article:
- The Privacy Grade rating (A-F) when visiting a website.
- Scoring of websites in terms of service and privacy, by partnering with Terms of Service Didn't Read.
I'm not sure how these services can be provided by DDG and its partners without them knowing which websites I'm browsing to.
IMO, it's none of their business, even if they promise not to keep the data or give/sell it to anyone -- a promise which is not verified by any trusted third party, as far as I know.
All of this calculation happens on your device, and is not handed over to us. Additionally, our privacy policy is to not store or collect any personal information at all: https://duckduckgo.com/privacy
Our privacy policy is extremely strict, and is the basis for our company. If we are to be found violating it at all, we would be ruined, and would probably be criminally liable.
In addition, a lot of what we do, including this new app and extension is open source so the code can be examined.
We're working on that. Because the content blocking mechanism doesn't offer fine-grained controls, we couldn't make it seamless and still block the things we think should be blocked -- like Google Analytics.
Hmm, can't speak for other people, but I seldom whitelist anything. I just leave or avoid the site. I also has Safari configured to go into Reader Mode automatically. Makes for a very clean web.
Have DDG written anything about how you make money?
I wouldn't mind paying a modest yearly fee. And I also wouldn't mind getting to vote on feature-requests such as "Better support for presenting stackoverflow results".
> Because the content blocking mechanism doesn't offer fine-grained controls, we couldn't make it seamless and still block the things we think should be blocked -- like Google Analytics.
Just curious, what about the Content Blocking API makes this impossible? I was under the impression that you could block JavaScript from the rule list.
Some google analytics code is embedded into buttons that are essential for the Web site to operate, so if you block them then the website will cease functioning properly. To get around that, we still block it, but insert some inert code in its place that will make the button function again properly. This nuanced replacement is not available through the API -- it's more all or nothing.
What is the advantage of this over Firefox Focus? And is the Privacy Essentials extension available for Firefox for Android? Trying to understand why I should use the DDG app over either of those two, since the DDG app seems like it is just a browser without tab functionality.
Also, not really related, but does anybody know how to make the home screen search bar on Nexus/Pixel phones use DDG instead of Google?
Tab functionality should be coming shortly, along with other standard browser functionality -- feel free to influence the roadmap with more suggestions :).
It is not available yet for Firefox for Android, though we would like to make it so shortly.
Our app additionally includes a few things:
--Upgraded encryption, in that if we can determine a site works on an encrypted version, we will also send you there automatically.
--Privacy grading, in that we show you at a glance how protected you are, which you can tap to dig into the details. We have more to do, though we've spent a long time trying to make this UX intuitive and informative, including an A-F grading scheme, grouping trackers by networks, and showing you over time who we caught trying to track you.
--Special attention given to blocking while at the same time not breaking the Web.
I've been using DuckDuckGo for a few months since I switched to the new Firefox. It's generally OK but I can tell that the results are just ever so slightly worse than Google. I get irrelevant results and old results. It's just obviously not nearly as smart and that's probably because it doesn't know as much about me as Google does.
DDG is usually good enough but I do find myself dropping back to Google for some searches (easy enough with !g).
It does make clear how tailored Google searches are to the user. For example, when I search for programming terms that have non-programming meanings (library names etc) I get a much more diverse list. Google pretty much exclusively only shows me programming-related results. With DDG I need to add my location to the search rather than relying on Google knowing my exact address.
I do find DDG location search very annoying as clicking on a map result always tries to open it in the Maps application, which I don't want.
This gives my (personal) site a "B", but literally the only thing it "fails" is that it doesn't have a privacy policy listed at the (third party) site "ToS;DR".
My site is a personal site, doesn't have logins, and doesn't even have any Javascript, so I think the "B" is a little misleading.
(Side note: Amusingly, if you go to https://tosdr.org with the plugin installed, they also fail for apparently not having a privacy policy listed on their own site. Also it says in a very large font that their ratings are outdated, so I really don't know how useful this is.)
We don't feel we can give anything an "A" until TOSDR vets the privacy policy, and unfortunately they don't have a ton of sites vetted yet. However, we are committed to working with them to dramatically increase the number of sites rated.
We're trying to put all the privacy essentials we can make seamless -- tracker blocking, upgraded encryption, private search and more to come -- all in one package, across all major browsers and platforms. In this respect, on any major mobile device or desktop browser, you should be able to look up DuckDuckGo and with one download get seamless privacy protection as you search and browse the web.
With regards to tracker blocking in particular, there are more nuanced UX differences. In the UI we're trying to move away from 'x trackers blocked' and instead group trackers from networks together, trying to identify the umbrella company and purposes.
Second, we're grading each site based on its privacy measures (including privacy policies with help from TOSDR), and telling you at a glance how protected you are on an A-F scale, based on what we could do (e.g. block trackers and upgrade encryption).
Ghostery makes/made its money selling things like a report on what advertising ends up being on your site, which you as site owner might quite reasonably want to know. (Advertising space is sold and resold at dizzying speed.)
IIRC Ghostery recently sold itself to Cliqz, which in turn is owned by Burda, a publishing conglomerate. Mozilla owns a bit too.
Yes, looks like they are now owned by some other company. Previously, one of many references [1] said this, which really smells fishy to me(Now not sure if the article itself is wrong):
"Ghostery is owned by Evidon, a company that collects and provides data to advertising companies. It has a feature called GhostRank that you can check to "support" them. The problem is, Ghostery blocks sites from gathering personal information on you—but Ghostrank will take note the ads you encounter and which ones you block, and sends that information back to advertisers so they can better formulate their ads to avoid being blocked. The data is anonymous, and Ghostery still does everything it promises to do to protect your privacy."
Seems logical - if you're an advertising company or an organized group of advertising companies and you know that people are going to be doing ad-blocking or at least tracker-blocking, if you can provide a tool that's 90% of the way there which also gets you metrics, etc. then at least you get that little bit of data and the existence of a "good enough" option may also delay stronger measures implemented by someone else.
That assumes that ghostery users block some ads, not all.
Try this alternative: The reports sold by ghostery tell example.com the link targets and reseller chains for each ad on its site, so it can keep away the ads for herbal viagra or fake gucci clothes. (Ad space is often sold through resellers, I've seen five-reseller chains myself and have heard stories about longer ones.)
Generally I am positive about this but I'd have to give some feedback that tag containers/managers are not automatically bad.
At the moment it seems like this tool just flags tag containers (e.g. tagcommander, google tag manager) as a "tracker network" which they are not. Of course, a tag container might contain tracking beacons and tags etc but they might also contain benign other scripts (e.g. for A/B testing).
I would prefer it if these sorts of tools blacklisted the tags inside the container (they are just loaded via ajax like everything else), rather than the containers themselves.
This is blocking Google Analytics, Facebook Connect and New Relic on my site. How would DuckDuckGo prefer that startups measure traffic, ad ROI and site performance? What's the "right" way?
The "right" way is to accept that privacy-conscious visitors reject your use of third-party tracking, and if you really need to track their visits you must use a first-party tool (Piwik, GoAccess, plain old logfile parsing...). It may not seem fair and it may be frustrating, but blocking third-party tracking is just so easy and anyone who cares about privacy can do it with minimal effort.
As a side benefit, when you start doing your own tracking from server logs, you gain valuable insight into where the gaps in data are from the third-party trackers, and precisely how many visitors are under counted.
You're right that third-party trackers aren't that reliable, and I do my own log analytics for the easy/important stuff, but I'm not about to roll my own New Relic.
This also seems bad for early-stage startups, where developer time is severely limited.
Early stage startups could simply skip all the analytics and do more qualitative UX research with smaller numbers of people they actually interact with in a human way and use large stats for things that actually matter like sales and revenue etc.
My biggest frustration with search engines is that certain websites I consider bad are common in results. These include sites that require a login to use properly (Pinterest and Quora) as well as websites that are known to have wrong information (W3Schools).
What I’d like to have is an option in my search engine (ideally that would not require a login) to always exclude certain domains from results.
That would improve my browsing experience more than any other feature. Blocking trackers is something others are already doing well.
I actually made that stories feed originally for myself, before DuckDuckGo even existed. As I still rely on it, I ported it personally to https://twitter.com/watrcoolr
To prevent suffering from tyranny of small decisions[1], I want to support duckduckgo, financially and technically. Do they take donation? Or do they have an open sets of challenge to work on? Would love to help.
I don't like the fact that the browser combines the search bar and the address bar.
It means that when I enter a URL character by character, the search engine will know where I am going next even thought I didn't need the assistance of a search engine in the first place.
A "privacy first" browser can do better in this regard.
This isn't really a big deal, but the features they are touting as new "today" have been in the FF version of the extension for at least the month or so I've had it installed. I'm honestly curious if I'm missing something or not.
We had to soft-launch early in Firefox because of the change in Firefox 57 of how extensions work, though were still getting out the kinks and preparing it for the rest of the platforms.
Kind of interesting seeing some of the ratings. Like Mozilla and HN have a B because of unknown privacy practices. Both these websites are fairly old, and it seems odd to me that the privacy practices of Mozilla are unknown.
How does the Safari extension play with the native content blockers like 1blocker? Is this supported? I noticed that with 1blocker enabled, the initial grade is usually higher than what it would be without 1blocker.
Maybe it's just a semantic argument because I am a search engine developer, but I wouldn't really call DDG a "search engine". It uses Bing to power results. I would describe it as a "search interface" that gives you advanced search related tools in addition to standard search, some stratified results, and then it takes a privacy angle to differentiate itself. As such, I think the move to browser safety is a natural extension of what they do.
That said, how the heck do they make money? I only very rarely see a single ad, and there isn't much money in NOT tracking people. It's very hard to make money on search if you aren't Google (I would know, I have tried several times).
Short answer is we're profitable since 2014 primarily with non-tracking search ads based on the keyword (type in car and get a car ad). While we don't serve that many ads, these are still lucrative enough to pay all the bills.
The affiliate portion is quite small now, and we do not change any rankings based on them.
One of the many nice features of DDG is that you can fall back to Google searches trivially enough by adding a "!g" at the front of your search results. The bang syntax also works for directing searches to YouTube, Wikipedia, etc.
I value my privacy, so the few times I have to redirect a search to Google is a small price to pay. Definitely recommend making the switch.
(Of course, if Google already knows everything about you, you'll probably find their results much better tuned for you, so YMMV. I was quite amused and pleased when all the suggested search completions for—IIRC—"git" in a recent DDG search were completely irrelevant to me.)
Choice is good. Big fan of Google and use a lot as super curious person but if does not work for you good there are other choices that work for your needs.
Can you please elaborate as to the issue -- I'm having trouble understanding exactly what is going on. If it is easier, you can email me directly at yegg at duckduckgo.com
OK, so when the Extension is installed in Chrome it takes over the omnibox search, directs it to DDG, and does not allow it to be changed.
Put differently, to get the features like tracker blocking, privacy policy scoring, etc. you must also be using DDG for search. I'd call that something that probably should be changed, or at least changeable in the options.
Unfortunately, this is currently the way Chrome works. We're looking for a way around it, but it looks like they changed this behavior sometime last year. In other browsers you can easily use another search engine without disabling the entire thing.
I occasionally get trolled around my previous company, Opobox, an early social networking company that I ran from 2003-2006. As far as I can tell, the trolling theme is that since I ever was associated with any kind of social networking company, that somehow taints me/DuckDuckGo. It's purpose was to help reunite old friends and classmates, and it actually helped develop my privacy views and practices for what became DuckDuckGo.
For example we did what I think were some innovative things:
--We collected the minimum amount of information possible.
--Allowed people to automatically remove all their minimal information permanently with one click or email.
--Actually charged money as a business model instead of using advertising.
--Never worked with any third-parties in terms of data targeting (and didn't need to since that wasn't our business model).
Of course, these also made the company ultimately not able to compete in the space, which was completely subsumed by Facebook.
I took these privacy ideas to DuckDuckGo, though, and realized in the case of Web search the minimum amount of information needed is actually zero. Hence, our privacy policy.
Unfortunately, like with social networking, there is an established freeness in the market, and so it is hard to charge people en masse. Fortunately for web search though, you can make plenty of money without tracking people, and so we've been profitable since 2014 and have no long-term viability issues.
w00t!! Is there some great public link that would be best to use in contacting the rare honorable advertiser who might advertise with DDG?
I'm thinking along the lines of a conversation about why it's unethical to feed Google / FB etc. given their fundamentally invasive business models and someone asks what to do otherwise, so I send them a great intro link that explains both that advertising on DDG is good ROI and more ethical…
"Install an arbitrary piece of code on your machine to improve your privacy."
Note that there is no link to the source code, no checksum. There's no indication that DDG isn't doing exactly what they claim to be protecting against. And the question comes up: what do they stand to gain from having users install a mobile app on their phone?
I'm not sure I understand what value DDG is providing exactly.
I use Bing with cookies and JS disabled. DDG on the other hand can't be used with JS disabled.
Before you say that privacy is about cookies and JS has no bearing in it, think that Google for example uses all kinds of JS to gather information about your browser. That's extra datapoints that they gather about you.
This announcement is more than private search; it's all the privacy essentials to search and browse the web -- tracker blocking, upgraded encryption, and private search (with more to come) -- all in one package, available on all major desktop browsers and mobile platforms.
With regards to the search component, we actually offer two Non-JS sites: https://duckduckgo.com/html & https://duckduckgo.com/lite. Also, turning off JS and cookies still doens't prevent a search engine from tracking you, which they can do simply by your IP address.
That's quite cool, thank you. For some reason it I can't make it work with Firefox's "search keyword" (the feature of starting a query with a keyword on the smart-bar which forwards that query to the search engine of your choice)
> Also, turning off JS and cookies still doens't prevent a search engine from tracking you, which they can do simply by your IP address.
Sure. Some datapoints about you are easier to hide than others. So are you saying that DDG doesn't store that?
>When you search at DuckDuckGo, we don't know who you are and there is no way to tie your searches together.
> When you access DuckDuckGo (or any Web site), your Web browser automatically sends information about your computer, e.g. your User agent and IP address.
> Because this information could be used to link you to your searches, we do not log (store) it at all. This is a very unusual practice, but we feel it is an important step to protect your privacy.
They also provide tools to help prevent other sites from tracking you:
>Finally, if you want to prevent sites from knowing you visited them at all, you can use a proxy like Tor. DuckDuckGo actually operates a Tor exit enclave, which means you can get end to end anonymous and encrypted searching using Tor & DDG together.
>You can enter !proxy domain into DuckDuckGo as well, and we will route you through a proxy, e.g. !proxy breadpig.com. This feature is part of our !bang syntax. Unfortunately, proxies can also be slow, and free proxies (like the one we use) are funded by arguably excessive advertising.
Can I give you a suggestion. When someone is using ddg with JS off, I think that dgg shouldn't "redirect to non-JS site". Instead you guys should have the same page being useable without JS using the <noscript> tag. Google and Bing both have achieved that, so it's possible.
I am not so sure why you are getting different results but the page redirects me to the '/html' section when I use the search-query and explicitly forbid everything (besides html).[0] And it is readable, even if css is disabled.
Yes, I agree that the JS-less version could be more discoverable. I learned about it years ago when DDG was new, in some online discussion forum - perhaps even here on HN - and have been using it via a keyworded bookmark ever since. I never even visit their title page.
I'm trying to get this working on Firefox right now. Doesn't work for some reason. I'm trying to associate keyword "d" with duckduckgo.com/html, but instead I just land on duckduckgo.com/html itself.
Interesting. I just tried using it on the tor browser (which comes with noscript) and you're right, it does redirect.
However when I access it on Firefox + umatrix with all js disabled, it doesn't.
I suspect I know what it is. Possibly noscript sends some header saying that the browser doesn't have JS which causes ddg to respond with the redirect, whereas umatrix just breaks the whole thing entirely.
EDIT: I just tried going to about:config and setting javascript.enabled to false, and indeed now I get redirected. This has significantly improved my view of ddg.
AFAIK it's the handling of <noscript> blocks in the HTML: if JS is on in the browser, the browser skips the block and extensions can't easily activate it after-the-fact.
> However when I access it on Firefox + umatrix with all js disabled, it doesn't.
Be sure you have the setting "Spoof <noscript> tags" enabled if you want noscript tags to be rendered when 1st-party scripts are blocked. It's a per-site setting, but you can enable it globally in the "Settings" pane in the dashboard.
I bet this is great for current events like “when is Star Wars 1 playing” when the person probably meant Star Wars 8. But sucks when researching items or searching for a paper or particular product.
I started using Duck Duck Go and it works great. The privacy is an added bonus.