Hi HN!<p>I'm the creator of Runnaroo. The title is a fun little rib at Neeva, which has been getting a lot of press recently.<p>I initially launched Runnaroo in a Show HN [0] at the end of February, and wanted to do a followup because the site has grown considerably in features and users over the last fews of months.<p>The core idea of Runnaroo revolves around a search engine of federated data sources to provide the most relevant and highest quality results. We are now over 50 Deep Search sources, and adding more weekly.<p>Some examples of queries that I believe provide better results than peer search engines.<p>react.js:
<a href="https://www.runnaroo.com/search?term=react.js" rel="nofollow">https://www.runnaroo.com/search?term=react.js</a><p>creatine effects research:
<a href="https://www.runnaroo.com/search?term=creatine+effects+research" rel="nofollow">https://www.runnaroo.com/search?term=creatine+effects+resear...</a><p>metallica tabs:
<a href="https://www.runnaroo.com/search?term=Metallica+tabs" rel="nofollow">https://www.runnaroo.com/search?term=Metallica+tabs</a><p>parkinson podcast:
<a href="https://www.runnaroo.com/search?term=parkinson+podcast" rel="nofollow">https://www.runnaroo.com/search?term=parkinson+podcast</a><p>bootstrap collapse link:
<a href="https://www.runnaroo.com/search?term=bootstrap+collapse+link" rel="nofollow">https://www.runnaroo.com/search?term=bootstrap+collapse+link</a><p>[0] <a href="https://news.ycombinator.com/item?id=22422604" rel="nofollow">https://news.ycombinator.com/item?id=22422604</a>
<i>Deep Searching: the inclusion of relevant results from other targeted search engines to deliver better results quicker.</i><p>That doesn't sound like "deep" searching to me. That sounds like search aggregation, like DogPile used to do a couple of decades ago.<p>To me, "deep searching" would mean the company has its own crawler that indexes the content that the other search engines ignore or discard because it hasn't been updated in the last six hours. The world is losing its knowledgebase because companies like Google only care about what's trending, not what's information.<p>I want a search engine that shows me all the things that Google has decided aren't important because they're not trendy. Show me the stale web. Show me things that are so good they don't need to be repackaged every six months. Show me hobby sites, reference sites, stores of knowledge that don't exist solely to play the SEO game. Show me things I can't get anywhere else.<p>I'll give Runnaroo a chance. Hopefully it doesn't disappoint. The world doesn't need another bubblegum search engine.
Looks nice, thanks for sharing.
I do find the three pieces of configuration to be inconsistent however. The first two have you flip a switch _on_ to turn something _off_. "Turn Off Quick Directs" and "Turn Off Deep Searching" would probably be better as "Quick Directs" and "Deep Searching" where the toggle defaults to the "on" state. They look very similar to iOS controls so I think the users' understanding of on/off state is already there.<p>As for the third option "Strict Search On", "On" seems redundant at best and misleading at worst. Misleading because I'm not sure if the text is going to change when I toggle the switch. Meaning I don't know if this is a static label telling me what the setting is or a dynamic label telling me what the current state is.<p>In summary, I would make the button text more consistent and change their default states:
[x] Quick Directs [x] Deep Searching [] Strict Search<p>As a side note, the controls are different on the main page vs the result pages. On the homepage they are switches while on the results pages they are check boxes.
This is great!<p>Just for test I searched for "python read command output without check_output" and Runaroo has more relevant search than Google.<p>Runaroo: <a href="https://www.runnaroo.com/search?term=python+read+command+output+without+check_output" rel="nofollow">https://www.runnaroo.com/search?term=python+read+command+out...</a><p>Google: <a href="https://www.google.com/search?client=ubuntu&channel=fs&q=python+read+command+output+without+check_output&ie=utf-8&oe=utf-8" rel="nofollow">https://www.google.com/search?client=ubuntu&channel=fs&q=pyt...</a>
Wow!<p>Thank you everyone for checking Runnaroo out. I shared it this morning mostly on a whim, and definitely did not prepare even close for the level of traffic being on the front page of Hacker News would bring. Runnaroo has been my solo side project for the last few months because I believed a better web search was possible, and I greatly appreciate all the constructive feedback.<p>I am bringing the site down for the next few hours to address some issues, but will have it up again soon. Apologies for any timeouts!
A couple of bits of feedback: 1. The searches I tried did return decent / good results, so well done! 2. The image search results page is not very useful / usable - there's no keyboard navigation or filtering (add "vector" as a filter and I'll be back regularly). 3. Dark mode would be nice. 4. What's your business model?
I've been collecting upstart search engines since DDG started ignoring search syntax instructions. It passes my first test:<p><a href="https://www.runnaroo.com/search?term=%22do+any+of+these+results+contain+this+text%22" rel="nofollow">https://www.runnaroo.com/search?term=%22do+any+of+these+resu...</a><p>...More accurately than DDG, which returns lots of results, all of which do not contain that text. Other searches so far have been far less swamped by topten trash as well. Thanks! I'll add this one to my list of search engines to try when I need accurate results.
I searched for "common" - <a href="https://www.runnaroo.com/search?term=common" rel="nofollow">https://www.runnaroo.com/search?term=common</a> and noticed the first result had a json payload as the description, which isn't visible on the site. Seems like a possible bug.
I do not know how you're doing it, but I just ran quite a few searches for obscure film photography and chemistry topics and the results were much better than DDG and less commercialized than Google. Very impressive!
Was very skeptical - every time I try some new search engine it just gives really bad results. Tried some quite esoteric things and was very surprised - also for non-English content. Trying this as my default search now! Congrats.<p>Edit: suggestions could be useful. For now, I'm editing the Firefox plugin to get suggestions from Google.
It's a bit weird that searching for "twitter" redirects me straight to <a href="https://twitter.com/" rel="nofollow">https://twitter.com/</a> - I understand the intent there but I'd rather get search results - I know how to navigate using my browser URL bar.
I was wondering about the business model for this:<p><i>> We may also add an affiliate code to some results returned that result in small commissions being paid back to Runnaroo if you visit or make purchases at those sites.</i><p>I don't have an opinion on this one way or another, but it's an interesting approach.
I think there's still a lot to be done in search engines when contextual and personalized results are concerned. I want my search engine to search not just the public internet, but also the niche stuff only I have access to. If I type "pizza", I want local pizzerias. If I type a name of a friend, I want to see their Facebook profile, as well as our conversations on Messenger, Slack and WhatsApp web. If I type "xxx crashing with code 608", where xxx is an internal service of some organization I work for, I want to see past Github issues, Slack conversations, Sentry reports and Jira tickets. The fact my search engine can't search my emails, conversations and private resources makes it much, much less useful than it could have been.<p>I genuinely don't understand why people dislike it so much when software processes information about them, for their own benefit. I use Google, Facebook etc. a lot, and I've never witnessed any bad consequences of being tracked. If not for that strange aversion, technology could be so, so much better.
A small feedback: You have a circle on the page that says "A better search engine. Learn why →". But only the tiny text is clickable. It would be great if that whole circle was clickable. Bigger link targets are always better.
Results are really good upon first glance.<p>Dark mode please! Just a quick UI tip: theres a reason that neither DDG, Google, or even HN don't have cards or separators for each result. I think the backgrounds of each result make them all look the same, and it's kind of impacting readability. My mind gets distracted by the rectangles and the shadows, and its just one more thing to sort through mentally.
IT-related search results seem super relevant. I wish other subjects would work equally well. E.g. supplement/condition-related search could also result in links to sites like selfhacked.com in the top (the actual articles there include references to scientific papers to back every particular statement) which Google actively hides from its results despite it's among my first go-tos.
I tried searching for<p>cmov gcc<p>(url: <a href="https://www.runnaroo.com/search?term=cmov+gcc" rel="nofollow">https://www.runnaroo.com/search?term=cmov+gcc</a>). Very impressive! Actually gives me stuff that's, I dunno, relevant.<p>Actually included both words without me quoting them - I wish DDG did that - you normally need to search for "gcc" "cmov".
Tried a few queries, and the direct redirect when I type in facebook is more than questionable. If Google did this I couldn't properly research news or facts about facebook. It is a really questionable practice.
That said, I would rather use this page than DDG, as long as I don't get lies about privacy stuffed in my face.
Runnaroo has become my default search engine since a couple of months.<p>1. Search results are more relevant. In comparison to google. I noticed a reduction in spammy results like slant, fake blogs, gitmemory, etc. There is still room for improvement.<p>2. Claims to respect users' privacy<p>3. Languages other than English are handled fine. In this regard runnaroo is usable for everyday searches, unlike DDG. Queries in a foreign language produce results of the same or lower quality than those from google.<p>I know the author is reading this, so here is some feedback.<p>- Make the search engine available for firefox mobile. I know there is an extension but it can't be installed on phones.<p>- Improve the behavior of the search box in mobile phones.<p>- Keep on fighting spammy results
Nice work! One bug/feature I noticed: Once focused on the image tab, if I revise my search, the image tab loses focus and the primary search results tab gains focus. It wasn’t the expected behavior for me.
I wish the about page had some information about the people who built this because I feel trust is increasingly important in my rubric for choosing what services I use.
I've been using it since this was posted here. I am ok with the search results, given the privacy I get back.. I do miss the google cards though... type EPL get all results for the Premiere League... type NBA get all games coming up... type movies and see what's coming out near me.
this reminds me of dogpile <a href="https://en.wikipedia.org/wiki/Dogpile" rel="nofollow">https://en.wikipedia.org/wiki/Dogpile</a><p>is everything old new again? have we hit some sort of cyclic point in internet search quality where we need this once more? dogpile was ended when pagerank showed up.
I wanted so hard to love DDG, but it failed so hard esp for developer questions (90% of my searches).... but DAMN this is what I wanted in a replacement. I've already replaced google, though native dark mode would be cool, but I can just use stylish for that ;)
Why does it use "<a href="https://en.wikipedia.org/?curid="" rel="nofollow">https://en.wikipedia.org/?curid="</a> style URL for general information from Wikipedia? Doesn't seem very convenient.
Why did you name it Runnaroo? Search engine are supposed to be easy to spell and type. During the last 3 days of my usage, I misspelt Runnaroo every 3rd time I opened it.
Howto "Show HN": post a project, ignore all best (technical) practices discussed on HN, update your site to "Unfortunately, we are temporarily taking the site down..."
Looks good and I'm going to try it as my default search engine for a bit. There are a handful of small UI issues that make me wish it were open source so I could fix them.
It’s easy to make a decent small search engine: nobody is working to game your algorithm. Vastly harder to do this as you scale to google size and popularity.
Search seems to be down due to the HN hug of death (as stated on the landing page). Looking forward to giving this a bash whenever it comes back up again.
why is it that all search engines still look like Google in the 90's. Can't the UI become fresher?<p>also I think SEO really ruined the search experience. do you have plans to use other ranking factors? Maybe social upvoting or so to improve ranking of actual good content as opposed to crappy listicles which is 50% of search results right now?
i like the way it orders the search results and not listing meaningless more pages of results.<p>One thing i found while searching images is that, i cant start a new image search , because it brings me back to the "all" tab.<p>ex: search harry potter<p>go to images , now i wanted to search for hermione images alone, but it takes me all back to full web results.
hohoho, wow. Tried quite a few searches and the results are very relevant. This is great.<p>If this is just piggybacking on Google API, can the business survive?<p>Another question is... how will you make money?
> Unfortunately, we are temporarily taking the site down while we perform some updates. Apologies to our existing users. We are working to get back up to be better than before.<p>Nice