On Jan 22, 2008, at 6:37 PM, Linus Torvalds wrote:
Ok, here's an interesting patch based on the current 'next' (since it veryintimately requires the new in-memory index format).What it does is to create a hash index of every single file added to theindex. Right now that hash index isn't actually used for much: Iimplemented a "cache_name_exists()" function that uses it to efficiently look up a filename in the index without having to do the O(logn) binarysearch, but quite frankly, that's not why this patch is interesting.No, the whole and only reason to create the hash of the filenames in theindex is that by modifying the hash function, you can fairly easily dothings like making it always hash equivalent names into the same bucket.
This is fantastic. Thank you very much for actually taking this issue seriously despite the mess I made on the list. This is exactly why I wanted to discuss on the lists instead of hacking away myself - there are very smart people on the list (like you) that already know how git works that can come up with ideas like this while I would still be trying to figure out where the index code is even stored.
-Kevin Ballard -- Kevin Ballard http://kevin.sb.org kevin@xxxxxx http://www.tildesoft.com
<<attachment: smime.p7s>>
