Hacker News Reader

Comment by temp231239 | Hacker News Reader

One think that I would like to see solved by indie search engines is the ability to break the search bubble and see whats usually hidden on 10th page of google search results. This is a serious problem in all big search engines.

One interesting way to solve search bubble problem is to have an option to filter out results from high traffic websites and blogs who invest heavily in CEO and pollute search results.

Having this filter will surely open a completely different world of information that's very hard to search.

View on HN

> have an option to filter out results from high traffic websites

That's what Million Short does https://millionshort.com/

The first thing I searched, it shows

Oops.. sorry about that!Engineers have been notified and are looking into the issue. If you see a captcha, be sure to fill it out and submit to continue your search

Edit: it looks like it doesn't support anything that is not in Latin letters at all.

Thanks, worked well for me and I found information I couldn’t find on ddg.

Good grief, this is astonishing. The other day I tried tracking down the names for various avatar species in VRChat. Surely someone had a list curated... but bing, google, and ddg all let me down with ancient wikis or completely unrelated sites, no matter my quoting or alternate search terms (animals, avatars, creatures, etc.).

Very first hit on millionshort: https://www.vrcarena.com/

Edit: Though, I'm still trying to track down a more comprehensive list. That one's missing Nanachis, which are rather common these days.

That's also the first hit on DDG for avatar species in VRChat

https://duckduckgo.com/?t=ffab&q=avatar+species+in+VRChat&ia...

I don't know what to say. Yesterday, for me, it was no where on the first few pages.

That's the second link on google for me for "avatar species in VRChat"

I don't know what to say. Yesterday, for me, it was no where on the first few pages.

fwiw, It's not on google's first few pages for me on that exact same search, so you aren't going crazy.

This thread does show up though. There are so few relevant results that I wouldn't be surprised if this thread alone is giving enough weight to vrcarena.com to already be rolling out to some people, hence the discrepancy in these comments.

Wow I didn't think Google went this far to personalize results...

Okay, this is crazy. This week I spent 5-10 minutes scrambling my brain to find this article [1], it was the first search result for "no more free sodas" (now, to be fair, it's now the first result for me on Google, too).

Thanks for bringing this onto my radar, I've added it to my pile of cool links!

1: https://steveblank.com/2009/12/21/the-elves-leave-middle-ear...

Thanks, this is super cool! I just used it to discover new music(search "electronics music albums" and removed the top million hits)

This needs to be front page.

I was quite impressed from what I’ve seen so far. The most frustrating thing about google how unreliable/irrelevant the results are when your query has anything to to with something that can be payed for, you get immediately spammed by low quality commercial bs, while ddg is practically useless for very specific searches I.e an error message in an app etc.

I just want a reliable way to block affiliate link harvesting blogs. They are the fucking bane of the internet.

That would be number one on my list too. I normally skip but it can be endless until you got to a result that looks authentic and not some needy click farm.

It's always a shame that W3Schools comes up before literally any other reference, notably MDN. I've always attributed it to their SEO efforts.

W3Schools and now TutorialsPoint.

While they're not bad per say, they;re never detailed enough of a reference. 99% of my webdev searches are now "mdn <topic>" as MDN is typically the most detailed resource on the topic (short of reading the exhaustingly verbose standards).

Of course now that I search "mdn <topic>" for everything I'm not training google's engine to match <topic> with the mdn resource, which helps cement w3schools as the top result for the topic

> !mdn <term>

is the greatest tool of ddg. Hundreds of search shortcuts available as "bangs"

If you're using Firefox you can also go to MDN, right click the search box and click "Add a Keyword for this Search...". It's (almost) like a DDG Bang but local to your browser (and for me it's a lot faster too).

It gets even more crazy once you realize Firefox is just saving a bookmark with a form submission template and you can make forms that "send" GET requests to data URIs and if you combine that with the keyword search you can run tiny web apps from your bookmarks using "commands" in your omnibar!

My default engine is already set to DDG anyway, so I can just !bang in the addressbar already.

I find the ddg search results load faster and are better indexed than the MDN built-in search, same goes for most sites that do their own indexing and are also indexed on big search engines.

In my experience the search on MDN sucks, and I usually get much better MDN results by searching for "mdn <searchterm>" on DDG than I get for "<searchterm>" on mdn.

I use an extension aptly called 'Remove W3Schools' - I'll give you one guess as to what it does.

Cuil did that, just in a surrealist way.

https://en.wikipedia.org/wiki/Cuil

http://cuiltheory.wikidot.com/what-is-cuil-theory

> This is a serious problem in all big search engines

I have never really understood why search engines do not let customers create blacklists of domains they do not want to see in the results. Are they really so cynical bastards that they understand that the crappy SEO results they cram down our throats are so good money to them that they can't afford us to hide them?

Google used to allow that like 15 years ago.

Yeah I think I remember being able to hide Yahoo Anwers in some native way. Would love to get rid of Pinterest from image searches.

From a UI perspective, how would this be implemented in a privacy friendly manner?

You mean without the search engine knowing what you are blacklisting?

Perhaps there is no good way, but in the case of Google, it seems like it wouldn’t change the privacy equation to have this not be private from them.

No, I mean, in what way can this feature be implemented and avoid sharing any PII (assuming the list itself is not PII, since that list could potentially be the same used by other users)? What about a text file hosted somewhere, and the only identifiable information is the cookie containing the link to that file?

Are you looking for a way to avoid linking an individual user’s fingerprints with a blacklist?

Yes, or rather: avoid that the blacklist becomes an individual user's fingerprints. Appreciate any suggestions on how to approach that.

Yes

Maybe I'm misunderstanding the problem. You can easily jump to page 10 of the results and scour from there if you'd like. They are paginated.

If the big search engines broke the supposed bubble and put page 10 results on page 1, well, there would exist the exact same "bubble."

This is so on point, thanks for that. Pollution is indeed the right word for it.

Seriously, this. Try searching for anything even remotely medical related. 100% garbage in every search engine.

Reply on news.ycombinator.com