If they had developed a good powerful code search (custom semantic engine for most used languages, complex queries, exact/fuzzy matches for syntax, use of history, etc) they could have become the primary way you interact with code you don't know yet.
As it stands now it's simply more efficient to clone and use plain old grep, it's really sad.
... but since it's allowed if the fork has more stars, I guess they're just trying to save compute cost. Seems like a bad place to scrimp to me.
It's in alpha stage so far and lots of useful features (like multi-line search) are missing but you can already try it out for Crates.io and Hackage: https://codesearch.aelve.com. Indexing all of npmjs is coming soon.
I used to go:
press start btn --> type "utor" --> press enter --> uTorrent.exe executes.
Why would they remove such a useful feature?
I'm super curious to what all these companies are using for search because they are so atrocious that tend to not even work in a basic substring search solution.
It boggles my mind why GitHub's search finds results across various commits, giving patchy results. It would be far better and likely easier for GitHub to just search HEAD. This frustrates me about GitHub every day.
[1] https://docs.microsoft.com/en-us/vsts/project/search/code-se...