> This is in contrast to google where searches return things that don't even contain your keywords.

This has become so frustrating as of late, especially when you search for something like "c++ map" and it decides to completely ignore the "c++" part of your search, leaving you with pages of actual maps. Slightly exaggerated example since I can't remember any concrete ones off the top of my head, but the actual situation happens way too often.

google's "must include" feature is one of their worst features ever. it nearly always removes the most unique, defining search term. it has been showing up with increasing frequency. i can't remember a time where it was actually useful or wanted. and if i didn't want the word in the search query, i wouldn't have put it in there in the first place!

I have the same problem.

I think this is a function of trying to make searches more effective for unsophisticated users at the expense of users who know how to be more precise.

Similar to Apple and GNOME, it ends up penalizing the expert users to cater to non-expert users.

This is why you need to have the regular interface for people who shouldn't be expected to be experts, but have a separate interface for people who know what they are doing, because supporting both types of users is often mutually exclusive.

I also learned recently that if you search for a quoted string in gmail and there are no matches it will just show a few partial matches, instead of just telling you there were no hits. The google behavior you describe got so annoying for me, along with the inability to copy the link in the results, I switched to DDG. It still provides results that don’t include one of my search terms all the time. It’s really annoying behavior.

I just did a Google search on “c++ map” and didn’t see the behavior you describe.

What if I told you Google search results are partially based on behaviour data, meaning the results you get for a search potentially differ from mine

It doesn't happen to me so it doesn't exist does not apply in this case

Use duck duck go, my lad.

Thanks for reminding me! Just set it up as the default search provider in my browser.

Google search doesn't really work for power users anymore and I don't see that changing for the better anytime soon.

Alternatively, if you can't do without the google search algorithm, use startpage. Doesn't track and bubble you, anonymises the queries.

Indeed. Every single result in the first "page" was about c++ maps.

Google fixed this specific case and a few others last year: https://blog.google/products/search/improvements-searching-s...

"Slightly exaggerated example since I can't remember any concrete ones off the top of my head"

I can confirm what they say is true, but probably not with this specific sentence. It's also really not useful while searching for phone numbers (which it changes completely without a second thought).

My favorite feature was 'near' queries. One could project n-dimensional spaces to any arbitrary lower dimension with the right semantic and spatial distances.

With Altavista, the limiting factor was the searcher. And my nostalgia that over-estimates the size of their search database.

> One could project n-dimensional spaces to any arbitrary lower dimension with the right semantic and spatial distances.

I'm incredulous. Doing this well for unstructured text would be worth a lot in many industries today, let alone the entire web. If they did this well, they'd still be alive and well in another form.

Near queries are easy on unstructured text. Here's the Elastic systax: https://www.elastic.co/guide/en/elasticsearch/reference/curr...

Maybe a smaller search corpus made it feasible, but this is an accurate description of AV's functionality. From http://jkorpela.fi/altavista/ , which is worth reading in its entirety: The operator NEAR ensures that both are within ten words of each other in the resulting documents. The operator NEAR binds less tightly than the operator NOT and associates to the left.

I also do not recall Altavista ever banning me with a CAPTCHA for "searching too hard".

Can probably blame dogpile for the evolution of that

You just caused me to have a flashback. Waiting for dogpile results was such a pain.

Never heard of "dogpile" before. Are you referring to dogpile.com search engine?

A week or so ago I was frantically searching for "hidden bird illusion", and could no longer find the animated gif on wikipedia, and also nothing related to it on DDG nor Google, until about an hour later...

I just tried "hidden bird illusion" and it gave the michaelbach.de page within first page...

Thanks for the recommendation if you guys were referring to dogpile.com it's always good to have a backup search engine...

yes was referring to dogpile.com

dogpile was/is a metasearch engine. Back before google days there was so much unindexed stuff out there each search engine could bring back different results. Going to this one site and searching would bring back results from all the popular search engines at one time.. querying sites like (altavista, lycos etc.)

I remember distinctly when someone introduced Google (ca 1999) to me as an Altavista alternative where the good results were actually on the first page. As in, with Altavista you needed to rummage through the first couple pages (luckily back then you only had a couple pages' worth of results) to find what you were looking for.

Maybe what we miss is not exactly Altavista but a mid-2000s Google, also with the low occurrence of spam and sponsored content that we had back then. How about duckduckgo, does it fit the bill?

Try selecting "Verbatim" in the tools menu - it can really improve your google search on technical topics.

Can't believe I didn't know that. Thanks a bunch!

Learned something. Thank you.

I hope someone somewhere makes a new version of Altavista search, if only for the developers. Searching via google is easy for basic errors but is a huge pain if you are slightly advanced. With the capabilities being extended (and more and more people writing code everyday), having a separate search engine dedicated to the developers could cut short the search times tremendously.

There is still http://symbolhound.com/ for search that doesn't ignore special characters but the index is probably not that big.

I was just asking about this Google "feature" on Twitter [1]. I can't imagine why anyone would want "must include" to mean anything but, and if it doesn't mean exactly that to Google, then they shouldn't say it.

> does anyone know why Google does this with search their results? what does the word “must” mean here, practically?

1. https://twitter.com/albedoa/status/1047502325005336576

The "must include" bit is a link. If you click it, you're telling Google that results must include that term (rather than may include, which is the default).

Oh man I used the shit out of altavista! I swear I could find anything on there no matter how obscure, it seemed like magic at the time.

Yeah, Altavista was the shizzle back in the day. I miss it, too.

IIRC, it didn't even have its own domain name for a long time, it was altavista.digital.com. There's another company I miss, having used Vaxen back in college.

First decent index of the web and designed to let you find what you want to find instead of things paying to be found

I wasn't coding when Altavista was around, but I totally agree that Google is terrible for searching code snippets etc.

Fortunately, there's SymbolHound [0], which is designed for just that! Check it out

[0] http://symbolhound.com/

I miss the graphical clustering representation.

If you searched for something like "rust", you would clearly see multiple clusters which would include a big chunk of stuff around "ferrous oxide" and a smaller cluster around the "programming language".

You could click on the specific cluster you wanted, and the results would rank/limit to just that.

I would kill for that back ...

So true. So, so true.