Live data from Hacker News

Million Short allows users to remove up to the top 1M sites from a search set

forbes.com

41–50 of 59 posts

Re: Million Short allows users to remove up to the top 1M sites from a search set

#41

I use https://addons.mozilla.org/en-US/firefox/addon/g-search-filt... for filtering Google. Can't live without it. It would be interesting to have a Github repo with some precompiled filters based on certain business domains. But for me at least, a couple of rules are enough to solve 99% of the problems.

Does it fetch next result page to fill removed items?

Re: Million Short allows users to remove up to the top 1M sites from a search set

#42

I use https://addons.mozilla.org/en-US/firefox/addon/g-search-filt... for filtering Google. Can't live without it. It would be interesting to have a Github repo with some precompiled filters based on certain business domains. But for me at least, a couple of rules are enough to solve 99% of the problems.

I'm using DuckDuckGo and I just realised I should install something similar, thanks! https://addons.mozilla.org/en-US/firefox/addon/ddg-hide-unwa... EDIT: You can actually achieve something similar with bookmarks, using keywords and '%s' The idea is: take a link to a search query string of a search engine, and replace the query part with '%s'. For example, take the following search query on DuckDuckGo: cute hedgehog…

Nifty trick! Is there a limit to the number of site: filters?

Look for an extension that highlights the good results. I find that more valuable than filtering the bad results.

Re: Million Short allows users to remove up to the top 1M sites from a search set

#43

I use https://addons.mozilla.org/en-US/firefox/addon/g-search-filt... for filtering Google. Can't live without it. It would be interesting to have a Github repo with some precompiled filters based on certain business domains. But for me at least, a couple of rules are enough to solve 99% of the problems.

Does it fetch next result page to fill removed items?

I'm not sure. Results I want are usually on the first page though, almost always highlighted in green. The "*.edu" highlight is particularly useful.

Re: Million Short allows users to remove up to the top 1M sites from a search set

#44
post #6

I’ve always wanted to see a search engine thar takes the top SEO tips and penalizes sites that use them. For example searching for how to grow a garden. I never want to see howtogrowagarden.com. I’d prefer to find more genuine l, non seo juiced advice.

An approach I've thought of in the past would be to prioritize sites with fewer scripts and smaller stylesheets. Lots of .edu pages fit the bill

Re: Million Short allows users to remove up to the top 1M sites from a search set

#45
post #20

Earlier quoted context omitted.

What would we do without the garbage content that is eHow and wikiHow?

I had a friend who had a whole list of dozens of banned sites configured for his google searches, which made Google way more useful. And then one day, Google decided to cut that feature out, as part of their drive to remove power user features across their products.

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

Similar extensions exist for firefox.

Re: Million Short allows users to remove up to the top 1M sites from a search set

#46

Earlier quoted context omitted.

I'm using DuckDuckGo and I just realised I should install something similar, thanks! https://addons.mozilla.org/en-US/firefox/addon/ddg-hide-unwa... EDIT: You can actually achieve something similar with bookmarks, using keywords and '%s' The idea is: take a link to a search query string of a search engine, and replace the query part with '%s'. For example, take the following search query on DuckDuckGo: cute hedgehog…

Nifty trick! Is there a limit to the number of site: filters? Look for an extension that highlights the good results. I find that more valuable than filtering the bad results.

I just tried, DDG seems to be a bit unreliable, especially when youtube is involved:

https://twitter.com/JobvdZwan/status/950331823229472769

As for Google, the limit to a query is 32 words, apparently: https://imgur.com/a/XW1Qa

... however, it also supports inurl:, so you can easily filter out sites with manu subdomains (say, pinterest.com, pinterest.co.uk, etcetera), just by using -inurl:pinterest

Re: Million Short allows users to remove up to the top 1M sites from a search set

#47
post #40
post #32

Earlier quoted context omitted.

For cooperate intranets Google is not a possibility. There xapian, elasticsearch or lucene are the best, with xapian dominating the backend and the Java stuff the frontends. For the others SEO optimizations are a real problem, yes. You can only try alternatives, like searx.to, bing or asking around.

> xapian, elasticsearch or lucene are the best, with xapian dominating the backend and the Java stuff the frontends Elasticsearch uses Lucene under the hood, in my experience Lucene dominates the actual indexing and searching, although I'm not familiar with xapian. By users, I reckon SharePoint search (FQL) is probably the biggest although it is way behind Lucene in features. Google actually used to sell bright yello…

With "xapian dominating" I meant the technical side. Of course lucene has more marketshare, because of the better elasticsearch frontend, and browser support. I.e highlighting and jumping to the results in word or PDF docs.

I wouldn't trust Google locally neither, and it's expensive.

SharePoint search is unfortunately used too often, yes.

Re: Million Short allows users to remove up to the top 1M sites from a search set

#48

Earlier quoted context omitted.

You can instruct Google not to include sites with a given word by prefixing the excluded words with a dash. For ex: > buy hoes -sex -porn

Yes, but the rest is not solved by that. Google has really let the search game slide for a while.

But the goal isn't good search. It's good enough search with Google's narrative (usually ads) filling in the space between.

Re: Million Short allows users to remove up to the top 1M sites from a search set

#50
post #9

Earlier quoted context omitted.

Problem is that SEO is tied to appearing relevant, so to penalize SEO is to penalize relevancy. Splitting SEO from relevancy is often a matter of making better search Searching for least-relevant can be pretty random: it's easy to point at the center of a circle, but the edge of the circle is not a point

A perfect system would penalize SEO (appearing relevant) and reward actual relevance. Of course, this is easier said than done.

This is exactly what the search engines have been attempting for the last 10 years.
Post reply on HN