Live data from Hacker News

Creating a search engine for fun and because Google sucks

vincents.dev

31–40 of 45 posts

Re: Creating a search engine for fun and because Google sucks

#31
post #28
post #17

Creating a thin, non-working wrapper around DuckDuckGo as a "hire me" ad, claiming it's a search engine and mentioning Google in the headline because it'll get more clicks.

Yes, I don’t know how that reached the frontpage. A failed 15 minute experiment from someone who has no clue what they are doing

I think it plays to the topics that are so often popularised here: the decline of Google Search and Rust™.

Re: Creating a search engine for fun and because Google sucks

#33
post #10
post #5

So your "search engine" uses Google and DuckDuckGo to get results? and you are creating it because "Google sucks"? Am I missing something here?

Kagi does that, is a paid subscription and HN does praise it a lot.

It's a little unfair to say that Kagi "does that" and thereby imply that they merely repackage search results from Google and Bing. Kagi uses an array of different search sources that includes Google, DuckDuckGo, Apple, Wikipedia, Wolfram Alpha, and others, alongside their own small web crawlers. Their mission is to present search results in a user centric way and so they try to surface results that are directly useful to the user rather than the advertisers (they don't have advertising partners as far as I can tell), and correspondingly their business model supports that approach rather than being in conflict with it. https://help.kagi.com/kagi/search-details/search-sources.htm...

Re: Creating a search engine for fun and because Google sucks

#34
post #6

Earlier quoted context omitted.

Google is fine unless you need something really specific, chatgpt is better at this point for a lot of queries because Google give you the most simple answer no matter how specific you got with your keywords. I still miss the ability to use "".

Quotes still work when there are good results. As far as I can tell all that has changed is a) dropping obviously bad duplicate results like Stack Overflow scrape dumps, and b) adding in more un-quoted results when there's a lack of good results for quoted. I personally haven't had any issues with quoting.

Last time I investigated (many years ago), quoted Google searches only work in some countries. I don't know why. I dug that fact up from a Google help page somewhere.

Re: Creating a search engine for fun and because Google sucks

#35
post #11

Earlier quoted context omitted.

Plugging https://search.marginalia.nu/

Yeah I ran that on like $5k worth of consumer hardware initially. The only caveat is that you want a lot of RAM, and since consumer machines are mostly limited to 128 GB, that's a bit of a hard limit on how much you can do. In terms of storage space and compute, it's relatively fine. I'm on a single ~$17k server now, probably utilizing about 20% of it's capacity.

Naive question: would lower-level language help here?

Re: Creating a search engine for fun and because Google sucks

#36

Duckduckgo is now particularly riddled by AI spam. The top 5 summaries for lots of searches now begin with some variation of "In the fast-moving world of...". Utter shit. It'd be so easy to filter I wonder why they/Microsoft don't bother. Oh, wait...

How much is DDG Bing based?

Re: Creating a search engine for fun and because Google sucks

#38
post #35

Earlier quoted context omitted.

Yeah I ran that on like $5k worth of consumer hardware initially. The only caveat is that you want a lot of RAM, and since consumer machines are mostly limited to 128 GB, that's a bit of a hard limit on how much you can do. In terms of storage space and compute, it's relatively fine. I'm on a single ~$17k server now, probably utilizing about 20% of it's capacity.

Naive question: would lower-level language help here?

The problem domain is sprawling and varied, so you ideally want a bit of everything. You definitely need to do a bunch of custom low level programming, as a lot of standard tools can't cope with the data volumes involved. Like I'm not even storing known URLs in a DBMS anymore, the table would be too slow to update.

High level language with the ability to make downcalls to a low level language is probably a sane compromise.

Re: Creating a search engine for fun and because Google sucks

#39

Earlier quoted context omitted.

Yeah I think they are trying to get a search-results-only view without the ads etc. Perhaps they would like startpage.com?

startpage does have ads, it's just somebody else than google tracking you.

You have just claimed that startpage is lying in their privacy policy.

Please provide evidence for such a claim.

https://www.startpage.com/en/privacy-policy?t=device

"Our search result pages may include a small number of clearly labeled "sponsored links," which generate revenue and cover our operational costs. Those links are retrieved from platforms such as Google AdSense. In order to enable the prevention of click fraud, some non-identifying system information is shared, but because we never share personal information or information that could uniquely identify you, the ads we display are not connected to any individual user."

Re: Creating a search engine for fun and because Google sucks

#40

Earlier quoted context omitted.

Quotes still work when there are good results. As far as I can tell all that has changed is a) dropping obviously bad duplicate results like Stack Overflow scrape dumps, and b) adding in more un-quoted results when there's a lack of good results for quoted. I personally haven't had any issues with quoting.

Last time I investigated (many years ago), quoted Google searches only work in some countries. I don't know why. I dug that fact up from a Google help page somewhere.

Was counties or languages?
Post reply on HN