I don't think I'm likely to find what Im looking for in the 0click box.
Every time I've used hnsearch, I've done so because I was looking for something VERY specific (that I think I remember from HN) and I usually need to go a page or two deep to find it (and/or I filter stories/comments and re-sort). HNSearch is a great tool and I use it a few times per week, but I don't think it fits the "0click" feature very well. Maybe it works better for other people though, and I'm the minority.
Out of curiosity, do you measure the performance of 0click box functions (including who clicks through for more info)? And have you ever dropped one because it underperforms?
Honestly, I prefer Google's implementation of site specific search better. I'd rather search with DuckDuckGo/Google's engine, because its better, rather than be redirected to a usually bad site search.
I saw a comment on a hn BitCoin article. The comment said that we've got 100 years of experience with [this type of monetary system], and we know how to break it, and then he listed a bunch of trading algorithms, including one (I'm pretty sure) named "sawtooth".
Amusingly, I found this thread when I was looking for the same thing. Then I dug the link out of my email (I had saved it there) and came here to tell you the result, only to see others had posted it.
I'd love to see some of the search engines who are catering to more sophisticated searchers just tell us the BOSS, or whatever cache they are licensing, API functions and then indicate which ones they are supporting.
For all we know several of these search engines are all accessing the same cache and they are simply choosing to focus on different options that the cache license offers and making different arbitrary CGI decisions like slash bang whatever.
It makes sense to hide details behind "magic" if the users are unsophisticated. But it seems like there's really little reason to do this for sophisticated searchers.
Maybe there is and I am just not seeing the competitive advantage or value addition.
At the same time, I feel like we may have different search engines all using the same licensed cache source and trying to differentiate themselves on multiple implementations of idiosyncratic ways to access the same cache API, instead of focusing on more basic factors, like speed and privacy.
Maybe searchers just want to access that licensed cache which is too expensive for them to subscribe to. Maybe it's not so important all the fancy things one can do with CGI. Maybe the fancy things are important. Maybe they just crave "features". I don't know.
But my guess is the BOSS cache is probably not that hard to work with and that most searchers could get their searches done easily enough, in simple fashion, if they had their own BOSS subscriptions, without the need for lots of customisations (which equate to the exotic features of these search engines).
This is one that I don;t really like, since the !hn hashbang already exists that will take a user to hnsearch.com