[hlau]

Really surprised to hear that you’re not getting the relevance out of Google. Are most of your searches in a specific domain that’s somehow not covered properly by Google? And is there a search engine that does a better job for you? I’ve tried using services like DDG and find myself falling back to Google more often than not.

[dredmorbius]

The problem is generally typical of online content as a whole given SEO gaming and commercialisation. DDG's bang searches and the ability to fluidly target searches to specific sites with less typing and fewer hops is a key differentiator.

The most relevant quality content tends to come from published rather than online sources, or by going direct to source.

The Web has been a mistake.

That said, Google's SERP page content, layout, tracking, and advertising all effectively drop relevance by a tremendous amount --- I've got to consciously filter out Google's own crap on top of the irrelevant web results returned.

DDG's cleaner presentation increases effective quality by a subjectively-assessed factor of 2--10.

Date-bounded search remains one of the very few reasons to favour GWS for a specific search, though even that is highly unreliable. Often what I want is a searchable archive from a given period, not a guestimate of a date-ranged search over the live Web.

Even in Google Books, date-ranged search results very often fail to return content from the requested period.

[_nalply]

> The Web has been a mistake.

In hindsight, what would have been better?

[dredmorbius]

That's a good question. I'm not sure I have a good answer.

I'm also not sure that djin can be rebottled. The history of media advances has been that they tend to progress and proceed, and human culture changes around them, they do rather less adapting to human culture.

(I've become aware in the past five years or so of the study of media and its impacts on society as a whole. Elizabeth Eisenstein's The Printing Press as an Agent of Change only hints at the full breadth, but is one of the major works on the topic. She draws heavily on Marshall McLuhan's The Gutenberg Galaxy (literally: we live in the universe Johannes Gutenberg created), and there are numerous others who explore this, notably Adam Curtis and Neil Postman. Again, the Web, algorithmic social media, and mobile computing each bring their own twist. Again, this isn't the first time media's transformed society. I'd argue that every advance, from speech on up, has. The changes can be tremendous and catastrophic --- to the previously existing order --- as with the printing press and the Reformation and Hundred Years War.)

One useful approach is to look at each of what were touted as the Web's strengths, and consider them from the perspective of "what could possibly go wrong". Several principles of the sociologist Robert K. Merton are helpful here: overt vs. covert functions and phenomena, unintended consequences, and possibly self-fullfilling prophecies.

It's not clear to me what boundaries can be established for the Web, or what the consequences of a failure to establish those might be. Either case the future appears bleak.

[rnhmjoj]

I almost never get useful results out of a web search, nowadays, unless the query is very specific (like looking up a website I forgot the domain of). This is not particular to google search, which in fact I don't use anymore.

Trying to find information given only fuzzy details almost never produce relevant links; anything remotely commercial, like trying to find a product reviews, film to watch, a store nearby, etc. produces tons of synthetic websites full of ads that magically match my query; specific technical information can result in low-effort blogspam or total rip-offs from other websites (stackexchange answers, other blogs, etc.); and the list goes on.

It seems to me the only actually interesting material is now found in forums, message boards, wikis and other kind of websites where users generate the content. Unfortunately searching these is far from handy because they aren't always indexed or have archaic interfaces or require a login. I think search engines in general, either by prioritising revenue or being tricked by spammers and CEO, are now blind to the real information contained in the web. I wish for a search engine that would only index a curated list of genuine websites based on a topic, but I don't think we'll ever have one because it's not profitable.

[mda]

"I almost never get useful results out of a web search"

This I find impossible to believe. So you basically don't get any useful results for 80-90% of your searches? I wish you could give some examples.

[rnhmjoj]

Ok, here's a few real example where I had only partial (but 100% correct) information had a hard time finding the right answer:

1. There's a shell (program) which feature a built-in file manager inspider by ranger, I forgot its name: try to find it. Answer: [1]

2. There's a particular gas that can (temporarily) kill a smartphone, but you forgot which. Find the article about this. Answer: [2]

3. There's a blog post (well-known if you're into networking) that argues IPv6 was meant to replace MAC addresses. Answer: [3]

[1]: https://elv.sh/

[2]: https://www.ifixit.com/News/11986/iphones-are-allergic-to-he...

[3]: https://apenwarr.ca/log/?m=201708#10

[mda]

1. Only finds it if something in the line of "programming language and interactive shell with built-in file manager" for less, it indeed misses.

2. "(this) gas causes smartphones to temporarily deactivate" second result

3 "blog post ipv6 was supposed to replace MAC adresses" 3rd result

In some cases some slight change in the wording changes the ranking drastically. However, I am not sure "The old good google" would find these at all honestly.

[alliao]

YES! I still remember the earlier days of Google, when it was not only returning results, the results were a delight to click through. I genuinely felt happy using it. Infoseek was good in a way that it let you search within searched results, so you could filter down. Alta Vista was definitely larger, but Google was pure magic. Not only relevant but digs up interesting and rewarding well researched information sitting around in a little corner of a web.