Hacker News new | past | comments | ask | show | jobs | submit login
Lesser-Known Search Engines That Are Worth Checking Out (256kilobytes.com)
182 points by August-Garcia 9 days ago | hide | past | web | favorite | 64 comments





I don't know if DuckDuckGo counts as "lesser known" but I find it incredibly useful.

The !wa bang command is great for weird conversions. (Ex: Sometimes watching mad men I might plug in "!wa 100 dollars 1960" and out pops an inflation adjusted figure)

!fake can check a given amazon product for fake reviews

Stuff like Wikipedia (!w), reddit (!r) etc.

Some people complain the results aren't as good as Google but I think they have either forgotten and/or never learned how to search. Since the engine has no PII on you, it can't handle ambiguity well. A good heuristic is if your query would give you a disambiguation page on Wikipedia, add a few terms.

I've been using it exclusively for a while, and the only time I really go back to Google is for their maps.


Every time ddg is mentioned on hn, it receives a lot of praise in the form I use this or that bang - in essence negating its usefulness as a search engine and making it a glorified shortcut app.

>Every time ddg is mentioned on hn, it receives a lot of praise in the form I use this or that bang - in essence negating its usefulness as a search engine and making it a glorified shortcut app.

A huge chunk of what people do on search engines is essentially using them to query databases. If I know which database my answer is in, why add an extra step?

It's actually a point in DDG's favor that they forgo a chance at ad impressions and allow the user to jump directly to what is wanted.

(Though to be fair I suspect a big chunk of their income comes from Amazon referrals, which will give them income whether you search on the site or use a bang command)


I'd counter that with the fact that DDG returns SO answer cards and various cheatsheets when searched (i.e. "tmux cheatsheet".) I think Google beats DDG in some ways (better news search results, for example.) But DDG is a much better programmer-focused search engine IMO, they just don't market themselves that way.

Back to your bang feature point, I really appreciate that DDG makes it easy to search Google instead by simply using !g. Tech companies don't always make it easy to use a competitor, and there's something to be said for DDG not creating barriers to using other search engines.


>Back to your bang feature point, I really appreciate that DDG makes it easy to search Google instead by simply using !g.

Or you can use !s to use StartPage, which uses Google's results to seed it's own and is more privacy respecting


I'd now days say google is mostly a shortcut app too. 90%+ of the time I'm really just loozing for data from a well known aggregator (wikipedia, reddit, stack overflow, rotten tomatoes, etc.)

Google basically enhances that in three ways:

* It takes away the choice/effort behind selecting a source.

* It combines multiple sources in an ok way

* It gives ok results for things that aren't really just shortcut queries


That's what I was thinking as well. I think I prefer to have those features and workflows one universal shortcut away in a launcher like Alfred: https://www.alfredapp.com/

Yeah I have some friends who say ddg is unusable for them. I really wonder what kind of exotic things they search for, I have been using DDG for a couple of years now and rarely need Google.

(!gi for their images I use sometimes)


I've repeatedly tried to use DDG for technical searches and it's often missing the results I can google. That would be fine except that it's often the only result with the answer and not duplicated in another result. Maybe adding !stackoverflow (or somesuch) might work but not if the result was a blog.

I wouldn't call my searches exotic but merely specific.


I feel this often enough to notice, but I think it comes from mostly having used google for so long. After learning the little tricks (even unconsciously) moving to any new search engine will leave you missing them.

I think this is exactly it. As someone who has been using solely DuckDuckGo for a couple years, including for technical searches (I'm a software developer), I find using Google noticeably frustrating because Google doesn't give me the results I'm looking for but DDG does.

I think I'm just used to the nuances of DDG where others (who currently use Google) are used to the nuances of Google. You get used to crafting search queries in a specific way, and having to change it is noticeable.


I tried to use it for search stuff in France, as an English speaker, and Google works way better especially when I’m using the wrong words

Prefer Yandex - faster than DDG, better results and you can easily search on Bing or Google from the same search result page if you don't find the result satisfactory (they provide links at the bottom of the results).

I used to abuse the !sh bang but it's been "deprecated".

Nowadays my favorite are

    !wbm : wayback machine (buggy but still useful)
    !cache : google cache (as a super crude text only view)

What did !sh point to? I can't find it in the list of bang commands.

PS: How do you generate monospaced font on HN?


Prepend two white spaces before the line you want to format as monospace.

  like this

  thanks for the tip friend :)

sh ~ scihub

thanks!

I wish DuckDuckGo had the minus operator to exclude keywords from the results. Otherwise it is my daily driver.


I have been using Raspberry pi for minimal computing, and ddg is the default one. Can't complain :)

There's also Wiby, a search engine for "oldschool", static, single-person-curated sites: https://wiby.me

I've copy-pasted this to HN previously, but it's worth posting again -- explanation from their About page:

"Search engines like Google are indispensable, able to find answers to all of your technical questions; but along the way, the fun of web surfing was lost. In the early days of the web, pages were made primarily by hobbyists, academics, and computer savvy people about subjects they were interested in. Later on, the web became saturated with commercial pages that overcrowded everything else. All the personalized websites are hidden among a pile of commercial pages. Google isn't great at finding those gems, its focus is on finding answers to technical questions, and it works well; but finding things you didn't know you wanted to know, which was the real joy of web surfing, no longer happens. In addition, many pages today are created using bloated scripts that add slick cosmetic features in order to mask the lack of content available on them. Those pages contribute to the blandness of today's web.

The Wiby search engine is building a web of pages as it was in the earlier days of the internet. In addition, Wiby helps vintage computers to continue browsing the web, as page results are more suitable for their performance."

EDIT: Apparently Wiby also has !g and !b for Google and Bing redirections.


Wow. Link of the day right here. Thanks.

My test search on all of these engines is "atari", so all I get here is atari fan pages, and blogs about retrocomputing and restoring old hardware and stuff, no atari.com, no wikipedia, no steam page for atari collection sales.


I wish it worked as intended. Here's a sample query I put:

https://wiby.me/?q=bag+of+word+model+matlab

See the results and decide.


As I understand the engine solely consists of user submitted sites. I wonder how many pages they currently have in their database.

This is amazing!

"Million Short" https://millionshort.com/search

It's a search engine that returns results but without the first 100 to first million popular results with logarithmic steps. It's a great way to step out the normal popularity filter bubble and find new things.


Google less pinterest and ebay is like a dream come true.

What a great feature idea. I wish I could configure some regex settings in [search engine] to force it to never show me certain domains.

Gotta plug shodan, the search engine for refrigerators (and any other connected device/service)

https://www.shodan.io


Searx[0] is also worth checking out. It's fully free/libre and easy to host yourself.

It uses other search engines, but so does at least DDG and Ecosia that are listed there.

[0]: https://searx.me/


I switched from DDG to Searx. Both are pretty good, but Searx is actually FOSS while the only reason I was using DDG is because it's not Google.

There are actually several Searx instances you can choose: https://github.com/asciimoo/searx/wiki/Searx-instances


You can also just run your own instance on your PC.

I've been using searx self-hosted for search exclusively[0] for a year or so now, and it's excellent on the general search side, but could definitely use some work on image search. I often get no/crazy/repeating results, and have to stay with mainline to prevent breakage. I'll maybe hit google directly once or twice a month, buy only ever for image search.

[0]: I also have it externally facing for use outside of my LAN, with forced SSL and an http simple auth to prevent unauthorized access/0day exploits, and every browser I've tried works fine with simple auth in front of the default search. Even iPhone browsers are fine


You host the client yourself, but it still queries searx's servers, doesn't it?

I'm not sure how that's an improvement over DDG.


No, because there is no such thing as servers belonging to Searx. Whoever wants to serve a Searx instance does so.

It queries Google and Wikipedia and the like. It's a metasearch engine, so doesn't crawl the internet on its own.

It's an improvement over DDG in that it's free software: You can't host DDG yourself.


Have two engines I'd like to eventually see:

1) An engine that removes all results from corporate entities. Essentially all pages would be from mostly independent entities. Obviously this would not be a primary search engine in and of itself since it's restricting a lot of content that could be useful, but at times I'd rather see what people are creating instead heavily SEO'd corporation #64728321. The description of 'Yippy' seemed promising here, but a search for 'space' quickly lowered my optimism.

2) A truly semantic search engine. It's amazing that e.g. Google was founded 20 years ago. And it was a major step forward in that searches for 'Abraham Lincoln' would no longer return hardcore porn. But since then we really haven't really improved much beyond that. Imagine a search for 'pages updated within the past 30 days about the launch of the crew dragon excluding large media and all social media results' would actually return what I'm looking for. Wolfram Alpha [1] is a very good proof of concept here, where the entire internet could effectively be a subset of all results.

[1] - https://www.wolframalpha.com/


I would love an engine that knows my blacklist. Stop showing me results from xyz.com!

If they want to go the extra step and identify sister corps, that works too!


Google once had this feature where you could blacklist domains. Unfortunately they removed it like 5 years ago :(

You can still do this with the "Personal Blocklist (by Google)" Chrome extension.

https://chrome.google.com/webstore/detail/personal-blocklist...


Google can sorta already do #2 and what you are asking for is very hard in it's entirety. Wolfram Alpha is a great prototype but it also shows how hard it is to do in the general case. When Wolfram works it is amazing, but it often doesn't work. The problem is that there is too much ambiguity for the current state of the art of NLP and ML. However I do think it is possible to do what you are asking if you are able to constrain the context to one or a few major topics.

Not mentioned, but there's SymbolHound when searching for programming language syntax: http://symbolhound.com/?q=%26%3Amethod

I have been lately surprisingly happy with www.qwant.com . I feel like I need to add !g much more rarely than with duckduckgo.


"Ecosia’s operation leaves a negative CO2 footprint on the world – their servers operate on 100% renewable energy and each tree that they plant removes 1 KG of CO2 from the atmosphere." But then later it says that Ecosia search results are based on Bing? If they are effectively searching Bing in the Background, wouldn't they have to factor in the CO2 footprint of a Bing search?

https://www.givero.com - just launched. Givero is a search engine that shares its revenue with good causes you choose.

We silently opened to the public about a month ago, and we are currently putting the final touches on the site and story, and working on onboarding more charities. Right now you can support charities working for climate, animals and children.

Since we have a lot of DuckDuckGo fans on here, I think I should mention that Givero: * Has DuckDuckGo compatible bangs * Has Instant Answers (just launched the first 3, more to come) * Is Privacy centric. * Is Bing based, like DDG. * Is Euro hosted.

So a good mix of DuckDuckGo and Ecosia, with the key difference being that we donate 50% of our entire revenue to charities you choose.

Full disclosure: I am the founder. We're a small team based in Denmark, formerly working on Findx (privacy search engine with own index), which shut down last year.


How privacy centric are you guys?

We do not keep a history of your searches and has no user profile.

We do not use third-party analytics tools, so your IP is not shared that way.

Your IP is anonymized immediately in our analytics tool (self-hosted Matomo), and we don't store your queries there.

We anonymize our raw weblogs, which are only used for debugging purposes, after 5 days.

We do have to pass on your IP to Microsoft at the moment, but you have the option to turn off personalized results (the "filter bubble", basically) on the search result page. We are working on getting permission to further tighten the privacy options here, but Bing requires a certain volume before they're willing to discuss it (several million searches/month) as we discussed with their VP of Search Partnerships in Europe. So we're on-par with Ecosia right now, but working to be on-par with DDG.


I asked this the other day. But I am asking again as I didn't get an answer. Suppose I want to make a small niche improvement on top of existing search engines. I want to them to provide first level results and then I will provide additional filters on top. For example, I would get the results from Bing for example, and then remove all websites which are ad supported.

How should I go forward with this plan? Are there any APIs available? Or should I put on a biz dev hat and talk to Bing guys?



"Ecosia’s operation leaves a negative CO2 footprint on the world – their servers operate on 100% renewable energy and each tree that they plant removes 1 KG of CO2 from the atmosphere.?

Only 1kg? Most of a tree's "dry" mass is carbon, it gets that carbon from the atmosphere, and most trees weigh more than 1kg...


I actually went ahead and looked it up for you: "And that’s not all. Since we use our profits to plant trees, every search with Ecosia actually removes approximately 1 kg of CO2 from the atmosphere. How? On average, it takes around 50 searches to finance the planting of a new tree. An average tree planted by Ecosia will remove around 50 kg of CO2 from the air during its lifetime."

Yeah, but after the tree dies and while it decomposes, AFAIK, most of that carbon goes right back into the air. So replanting trees doesn't really have CO2-negative effect.

Having more forest area keeps more carbon down as long as the forest lives. However, if you want to actually get rid of the carbon permanently, you'll have to bury it.

Maybe 1kg per tree reflects some kind of average of these 2?

But I'm not an expert in this topic (not even very knowledgeable) so someone please confirm or correct me!


I assume these are not trees that are planted and then left to grow forever, but probably the trees are harvested at some point. Go ahead and write them an Email, asking how they got that 1kg figure.

Is there ever going to be search engine that will replace Google as Google replaced Altavista ?

If somebody created true AI, they could create a search engine with it that is better than Google.

True AI likely wouldn't be able to deduce context much better from the words given (i.e. a clear 'this' or 'that' situation, like Hamilton the musical or Hamilton the individual). It's just impossible to tell sometimes, and would require user effort to specify. Maybe the value in the AI could be sorting results into the 'this' category or 'that' category, making perfect results just one step away for the user?

Also, I've heard that the issue with AI in some specific domains is substantially slower responses than traditional / non-AI. Would this / how would this be reflected in a search engine? Would query results be updated less often?


True AI would perhaps know the user well enough to know their intent (even Google already knows the user and perhaps even know they have a likelihood for a person being more interested in musicals or in maths).

It could also actually understand web sites and create meaningful summaries. And all sorts of other things.

As for speed, I don't know - it seems in many applications AI is way faster than traditional approaches. Also things can be parallelized. DeepMind trained for many years (measured in computing time), but it was parallelized so it didn't actually take years to train.


> At this time, Blackle claims to have saved over 7 million watt-hours of energy.

That's only 7000 kwh, or 100 full charges of a Tesla. Seems more like a gimmick than anything impactful.


> Learn about a handful of helpful search engines that are not Google, Bing, or any of the other really obvious ones

The engines behind all of the general-purpose search products mentioned in the article are still Google or Bing.

I would love to see a competitor.


Can websites please stop using these bright, loud colors to display text and html links? Its very hard for me to read this article!

I can barely read orange over blue text...

Nice. I'd forgotten Wolfram|Alpha.



Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: