Krugle - A Search Engine for Open Source Code

Posted on June 5, 2007 by Guy Snir
Filed Under Open Source, Search |

Krugle is a Search Engine that positions itself as the place for developers to search for code, documentation, bugs, etc.

The motivation behind Krugle is that most developers today use a search driven approach, where development begins first with searching for information (which is a rather safe statement).

I like the filtering capabilities (Language, Found in, Project, etc.) but I think (based on some simple attempts) that the search algorithm could still be improved (I guess I am spoiled by Google…)

Some recent articles on the subject are at: Linux.com, searchenginewatch.com

Comments

2 Responses to “Krugle - A Search Engine for Open Source Code”

Ken Krugler on June 5th, 2007 9:32 pm

Hi Guy,

Thanks for taking a look, and giving Krugle a try.

I’d love to find out more details about what types of searches you tried, where you didn’t get the results you were hoping for. Was it that we didn’t find hits in projects you know about, or the ranking of the hits wasn’t what you wanted, or something else?

Thanks,

– Ken

PS - One common issue is doing “semantic” searches in code - I’d blogged about this a few times in the past. See http://blog.krugle.com/?p=227 for more details.
Guy Snir on June 6th, 2007 2:22 pm

Hi Ken,

Thank you for your input. I tried to recall the searches I tried in the past but couldn’t…

I tried some new searches, and I think that your post about the “semantic” searches answers most of my questions (funny enough I tried a search for “parse xml” before reading the post!).

Now that I tried some different types of searches (i.e. “reverse polish notation”) there were great results.

Bottom line, I stand corrected, and I added Krugle to my Bookmarks…

Thanks,
Guy