Ask HN: What should the next web search engine be like | Hacker News Reader

Ask HN: What should the next web search engine be like

4 pointskreeben3 comments

It's hard for us web search engine builders to ever become finished with our MVP because you guys keep moving the goal post.

First you were like, hey, stop with this semantic search, it sucks. Then you were like, hello, verbatim by default, yo. Then you were like hey, don't track me, bruh. Then, what's up with all these ads, bruh? Then you complained about poor search results. Then reviews become very, very important. Who can you trust? Then you wanted an SEO proofed web search engine. What's canon? Then you realized you wanted to be able to exclude certain domains from your results, because you know who to trust. Also, it must be FOSS!

Is there anything else that I've missed? Me and probably also all of my fellow web search engine builders out there who do this backed with money or as a side show would sure like to know.

If we can decide in scope, please, that'd be great. Then you'll start to see super focused contenders popping up on Show HN in no time that mean business. Let's stop complaining and let's do something about search!

Scope freeze after this meeting.

TL;DR: Google used to reach into the World Wide Web and deliver what I asked for, now it sells me what it has on hand in its stockroom.

If I type terms that map unambiguously to a result into 1998-201x Google, I get a well-regarded exemplar of that result as the first result, followed by less-trafficked exemplars of the same result, followed by things that are related to the result, followed finally by irrelevant results. It was so consistently good that I took it for granted for 15-20 years.

If I do the same today, most of the time I get no relevant results on the first two pages. I get fresh results from popular or well-funded sources that happen to have a few of the search words, but are otherwise 100% irrelevant. It's like asking a shopkeeper for an orange, and they say "You're in luck! I bought way too much orange paint so I have tons of that in inventory. What do you say? Eh?"

I always want the former. I never want the latter. I don't know how I can be any plainer. If that isn't specific enough for a would-be contender, then frankly that company doesn't belong in this space at all. If you think I'm looking at the past through rose-tinted lenses - that's fine, that's what they all say - but your search engine is still going to fail.

Honestly, to me, I'll be happy with a good FOSS search engine.

The only thing out there is gigablast but it needs money/ hardware/resources and a better look GUI.

I had high hopes about brave... But they are not FOSS so... Nothing new there. Just for privacy I prefer DDG (even depending heavily on bing and that I would prefer it was independent - I don't like much of that meta mambo jumbo)

All in all: an independent free and open source search engine that respects user privacy.

I don't think it's too much to ask

Broad ask, but I have three things on my mind:

- a good time control allowing results filtered by specific time periods

- a type of content filter set to include/exclude content by type of site: social media, news, personal sites, blogs, etc

- show the last crawl time in the results (per result item)

Reply on news.ycombinator.com