New Search Engine "Brave Search" Beta Launches to Public

First new search engine in ages!

[engine]

Brave Browser has launched its search in beta using its own database.

Its principles include;-

Privacy: no tracking or profiling of users.
User-first: the user comes first, not the advertising and data industries.
Independence: Brave has its own search index for answering common queries privately without reliance on other providers.
Choice: soon, options for ad-free paid search and ad-supported search.
Transparency: no secret methods or algorithms to bias results, and soon, community-curated open ranking models to ensure diversity and prevent algorithmic biases and outright censorship.
Seamlessness: best-in-class integration between the browser and search without compromising privacy, from personalization to instant results as the user types.
Openness: Brave Search will soon be available to power other search engines.

[brave.com...]

Earlier story

Brave Search Coming to Privacy-Preserving Brave Browser [webmasterworld.com]

[Dimitri]

Brave has its own search index

I can't find information about how they are building their index, and how to identify their crawler.

[jmccormac]

That user based ranking was tried with Jimbo Wale's Wikia search back in 2007/2008 or so. It didn't work back then due to a poor index. Be interesting to see how they've solved the problems.

Regards...jmcc

[iamlost]

Brave Search beta is based on Tailcat (open source SE from Cliqz that shutdown last year), which they acquired this past March.

Brave needs to acknowledge their crawler (I assume they have one but...Cliqz had Cliqzbot), IPs crawled from and/or rDNS if they want to keep their reputation as squeaky clean private.

Currently they utilise Cliqz-Tailcat index with (optional?) anonymous ‘mixing’ of Bing/Google results where own index is insufficient (displaying percentage of third party returns).

[iamlost]

Added: I’ve been told that Brave is extending its Tailcat index off the queries and clicks of the browser - an advantage of having the SE integrated with browser and mixing in third party results.

So, apparently no crawler. Interesting.

[jmccormac]

The index might still be an issue if some of those domain names are new registrations from the last few years. The current non-renewal rate for .COM is around 44%. Without crawling that index will get stale. What killed Wikia Search (apart from trying to get people to develop a search engine for free) was index spam.

Regards...jmcc

[edited by: jmccormac at 7:24 pm (utc) on Jun 23, 2021]

[NickMNS]

I wonder whether they will include results from the IPFS protocol, ie from .eth TLDs

[ecommerceprofit]

[search.brave.com...] or search from the address bar

I actually love it. Better than duck.com

[Brett_Tabke]

I think it is a good start for them. Extremely interested in this one.
Can't wait to hear more about their tech.
Only concern is they say they 'back fill' with results from Google when they don't have enough results.

[RedBar]

It is nice and clean and quick but some results are simply incorrect / out-of-date however saying that I have now added it to my list of engines to use, I'll definitely keep checking it out.

[RedBar]

For the past hour I have been trying to find information about a specific sport in my region to no avail and actually sent out a load of texts to friends asking if they had any ideas?

To my surprise Brave just came up with precisely what I was looking for, I had a doh! moment when I realised the answer was so obvious, good job Brave.

Interestingly none of the other engines I tried came up with this obvious answer.

[Brett_Tabke]

So Brave acquired "Tailcat" last year. [itsfoss.com...]

Can "Tailcat" still be downloaded somewhere? (searched and searched)

[JS_Harris]

The political news bias is instantly visible. Since they don't crawl but they show stories 20 minutes old who's providing the stories?

[Dimitri]

I wonder if they (can) use their browser, to index on the fly, pages visited by the users.

[brotherhood of LAN]

A lot of the technical background of cliqz/tailcat can be read about here: [0x65.dev...]

> I wonder if they (can) use their browser, to index on the fly, pages visited by the users.

Part of what cliqz done with their forked FF browser was send telemetry about searches you've done on Google and which results you click on, and then crawl those pages. Some of it is covered here: [0x65.dev...]

[ecommerceprofit]

>Only concern is they say they 'back fill' with results from Google when they don't have enough results.

Agreed. Hopefully this will be temporary and they will continue to build up their own index. Eventually they need to cut the cord and stand on their own.

[ronin]

iamlost wrote:

Added: I’ve been told that Brave is extending its Tailcat index off the queries and clicks of the browser - an advantage of having the SE integrated with browser and mixing in third party results.

So, apparently no crawler. Interesting.

Dimitri wrote:

I wonder if they (can) use their browser, to index on the fly, pages visited by the users.

If I've understood this correctly, this is really clever marketing for the Brave browser, if it means that people who want webpages and websites (and not necessarily their own) to climb the SERPs can influence that climb by browsing the web using Brave.

Those working at the SEO coalface will switch from Chrome to Brave - because there's really no downside.

Those not working at the SEO coalface will be increasingly advised / recommended to use Brave instead of Chrome by those who are.

Any initial reservations will be easily waved away because... Brave uses Chromium / Blink / V8. Just like Chrome.

To the average user, the two browsers are near-indistinguishable.

1990s: Netscape Navigator
2000s: Internet Explorer
2010s: Chrome
2020s: Brave ?

[Dimitri]

Chinese crytpo-famers, who recently lost their job, can find another one, spending the day, browsing web sites with Brave, to make their ranking improve.

[JS_Harris]

I wonder if they (can) use their browser, to index on the fly, pages visited by the users.

Once spammers realize they aren't getting the content themselves directly it wouldn't take long before the stream is poisoned with spam. It would also mean your browsing habbits are not private.(hey, we aren't watching you, but we're recording everything you look at and sending ourselves that data from your computer, hmmm)

As of right now Brave search has the same bias as Google search making it the same animal in a different skin. When an unbiased search engine comes along people will instinctively know it. I think it's pointless to declare yourself anything anymore, with trust so low whatever you claim to be must be shown, the skin has little to no effect. I'm not seeing what's being promised, yet.

[Manish Bhickta09]

Good Luck To Team!

Anyone knows in which market they are focussing?

Do you guys believe, Brave is a direct competitor for Google in a decade?

Any thoughts/views

[Achernar]

In the article ''9 Features in Brave Search" https://itsfoss.com/brave-search-features/ (thanks Brett_Tabke) we can see examples of options: local / global, (settings) dark mode & tweaks, "find elsewhere". I don't see any of these.
Anyone seeing them? Is it only enabled inside brave browser?

[Brett_Tabke]

Say, after using this extensively this week - Brave Search looks like the real deal. Extremely high quality results and still in BETA!

[Brett_Tabke]

and how about this throw back from Brave Search:

Find elsewhere - Google - Bing - Mojeek

actual links to other search engines.

Mojeek? Still around. wow.

[jmccormac]

Mojeek is still around. Have seen it crawling in the last few weeks.

Regards...jmcc

[heisje]

Sorry, that's beyond hopeless : laughable. Index miniscule, a joke.

What is exactly the point? Competing against AltaVista?
That one has been dead & buried long time ago.

.

[NickMNS]

To answer my own question.
It appears that the Brave search engine doesn't support or search the IPFS protocols sites ie: .eth domains. This despite the fact that the Brave browser is one of the first to support the protocol. See here:
[brave.com...]

I really think that this is a major missed opportunity. If Brave wants to position itself as a Google Buster, then providing search functionalities for IPFS would be a major feature. Without something to differentiate itself, Brave is just another DuckDuck novelty that sends no traffic.

And yes I know, that most people here are thinking "IPFS what's that?" and "Who cares!", but with the current rise in popularity of crypto-currency and blockchain technologies, this as an innovation that needs to be watched.

And for those thinking along the lines of:

The political news bias is instantly visible.

The only way to truly avoid a bias is to have a truly decentralized internet. Where governments can't revoke domains or block content.

[ronin]

@nickMNS

Good points regarding IPFS.

I'd also like to see mainstream search engine and mainstream browser support for Hypercore (formerly Dat).

Brave Browser and Brave Search might be good candidates for this.

[theregister.com...]

[blog.datproject.org...]

[hypercore-protocol.org...]

[thor_swe]

Brave has been my default search engine for 3 days and it has performed well over my expectations.

I received useful results for ~85% of my queries, that's pretty darn good for a SE in beta with it's own index. Just sayin'.

[JS_Harris]

Code

Brave is built on Chromium, Google's open source code.

Fallback

'for queries where Brave results are not yet refined your browser will anonymously check Google for the same query...'

Odd choices for a 'private Google alternative', imo.

[ronin]

That's a little unfair.

If anyone wanted a privacy-emphasising alternative browser which wasn't chromium-based, they'd use Firefox (or Firefox-based Tor Browser).

If anyone wanted a privacy-emphasising search engine which didn't rely at all on Google, they'd use DuckDuckGo. (Which relies substantially - or entirely? - on Bing.)

Brave fills a gap by providing a privacy-emphasising alternative to both Google Chrome / Google *and* Firefox / Tor Browser / DuckDuckGo.