Ask HN: If you hate Google 2022, why not making your Google 2012?

6 points by frozencell ↗ HN
Most of the source code of Google 2022 seem to be on Github separated (PageRank, MapReduce, Freebase, Graphd, etc.). Why isn't there no unified open source 2012 Google-like search engine?

11 comments

[ 5.2 ms ] story [ 22.3 ms ] thread
I don’t hate google but Google is not just search, it’s an ecosystem. OS, browser, email, search etc.

Wrt to Search,

Scale: Huge cost of infrastructure, data is privatize.

You could run a hobby scale pagerank but for searching web it would be a needle in a haystack.

They have competition like yandex, nacy, baidu but are not for English web and not open source

In 2022, Android & Chrome are the most valuable pieces of Google.

For instance, if Microsoft controlled them, they could just point the search functionality toward Bing, and most users mightn't even notice.

Why following the same steps in the wrong direction?

When future search engines begin receiving feedback from users and ranking accordingly Google will become irrelevant.

To compete with Google you can’t just use their legacy tech (pagerank, mapreduce, etc.), you have to come with a better tech.
If the legacy code runs faster on modern hardware then the modern code, wouldn't that be a competitive advantage?
There are probably much better optimized algos nowadays, and new tech will require a different way to store and retrieve data.
I don’t have the resource to run a 2012 search engine. Even with the CommonCrawl dataset, it’s extremely expensive and also not environmental friendly.
Running indexes and crawlers is expensive and google has monopoly power on monetizing them
Can we give up our CPU during off peak hours to contribute to crawling? I would do this in return for a Google 2012.
I think that the next search engine is gonna be a service that has an advantage using a technique to identify quality content and ditching the rest, most of the internet is just noise.