Ask HN: Open Source search engine proposal. Would you pay for data files?
elefound.com
elefound.com
I suspect they'll be large though, and require actual infrastructure to support. Probably a mid-sized VPS at minimum.
At that point, I don't think I'd want to buy them anymore. It's going to cost more than the new crop of premium paid search engines, and I still have to do the maintenance myself.
It makes more sense to me as B2B operation than a B2C, as long as infra is required.
Lately I have been admiring the efforts to make the web less “corporatized” with the open, decentralized projects of the fediverse (but not all the useless blockchain stuff that’s usually associated with “decentralized”). I think this project is a necessary step towards the future of the web, but I can’t comment on the financial viability of it.
Would you be willing to pay for the creation of search index data files?
(But for people to have an opinion, I thought that some context is needed)
Those can also be created by crowdsourcing: https://yacy.net.
But I believe that for kicking off such a project, it first needs to exceed a critical mass of search coverage on mainstream targets. That's why I am interested to see if it could be executed while being financially self-sufficient. Taking on VCs early would probably undermine the project, so bootstrapping seems like the way to go. And that means you need paying customers from day 1.
Relying on friendly volunteers, on the other hand, seems like quite the risky strategy.
EDIT: BTW, looking at their screenshots, it looks like Yacy is planning for a very different scale. I was actually considering indexing pages in the 10mio to 50mio URL range. But all their screenshots show <10K URLs, so that would be mostly niche results.
Those are just demonstrations. According to the live stats, >10mio URLs is totally possible: https://yacy.searchlab.eu/Network.html.
wouldn't some sort of group moderation be more effective?
The idea is that you have configuration files where you - the user - write down all the domains, companies, etc, that you want to get search results from. So if you never put that spammer's domain into your local config file, you will never see their results.
But since managing all those config files by yourself might be a lot of work, and because less technical people probably can't do it, you can alternatively subscribe to someone else's configuration file so that whatever they add to their config is added to your config automatically, too.