Hash
 

When will we finally see HASH implemented in LimeWire? Even Gnucleus supports HASH!!!!!!!!!

Well, the LimeWire team is working on it, so it will be released when it's ready, I guess. I assume it will come in 2.5 when we will probably see:

Hashes, full HTTP1.1 support and the return of browsing hosts.

Yes, well the LimeWire team has been promising this since version 2.2

But this time the code is already on their server...

Really? then what are they waiting for?

Testing and fixing bugs I assume. It's not been added to the cvs mainline yet, so I believe there are some issues left...

I wish instead of gnucleus "hashing in the background" I could just hit a button called "hash all unhashed files NOW!!"

Grr..


Edit: LimeWire should have that button, too (in the library window, or something)

currently the limewire cvs code creates the hashes on the first startup all at once (may take a while on systems with large/alot files).

Why don't they release a beta?

Most of the new features are still under development.

sure hash helps alot (downloading a file from 19 hosts right now) but then again, I am not finding as many files as I used to.

I rather doubt hash couldbe helping at all already. Most people don't have versions of LimeWire that hash. You can't hash files on other people's computers, so the vast majority of search results you get are not hashed yet.

Not to mention that LimeWire doesn't make use of hashes yet. - So far LimeWire won't recognize identical files with different file names, even if their hashes are the same. - If you want to change that, simply change the getFile() methods in IncompleteFileManager.java, so the filenames for incomplete files would be based upon their SHA1-hashes instead of their names and sizes.

- I tried that this weekend. It didn't help my download experience very much, since some 75% of all files aren't hashed yet. Once this ratio reverses (with the transition of Morpheus to a newer Gnucleus codebase, I hope this will happen soon) LimeWire could easily use hashes to verify two files are identical.

For the search view, you'll have to change the TableLine.java ( the matchInternal() method), but I found that grouping just files with identical hashes is too restrictive and everything else that could be done without completely rewriting the grouping mechanism didn't show much effect since most of the files simply aren't hashed yet.

AlternateLocations are also just collected yet but not actively used.

Where can I get the hashing version?

This is all in the new beta. We don't fully take advantage of it yet (we don't propagate the download mesh as much as we could) but the infrastructure is there to allow us to layer on some nice functionality.