Increasing the number of possible sources

[mls]

Apologies if this is already done, or has already been requested, or can't be done, but...

Would it be possible to integrate Bitzi or the Hash codes into the search itself, so that all identical files, regardless of the filename, were identified together?

If that's not clear, let's try a simple example:

If I search for "Antmusic" my search results might contain:

Antmusic.mp3
Antmusic_1980.mp3
Antmusic - Kings of the Wild Frontier.mp3

And so on. Now, each file may be hosted by a single user, but if all three files are identical, exept for the filename, why not download the file from all three hosts?

Better still, why not show it as a single search result with #=3, using the first or most common filename in the results list?

Thanks. MLS

[foolofthehill]

And how should the Gnutella network know that file XXXXX.mp3 is the same file as XXXXX_1980.mp3 ??

Guess that is not practicable..........
But that's just my 5-cent-worth.....

[Sleepless]

I think it allready is like this. I have on several occations tried to download something with only one source, and it goes up to 6 or 8 hosts right away

[mls]

Originally Posted by foolofthehill

And how should the Gnutella network know that file XXXXX.mp3 is the same file as XXXXX_1980.mp3 ??

I think the answer is related to how Bitzi works. As I understand it, every file has an (almost) unique 'hash' code, which allows you to identify it, regardless of its filename. MLS

[Sleepless]

Originally Posted by mls

I think the answer is related to how Bitzi works. As I understand it, every file has an (almost) unique 'hash' code, which allows you to identify it, regardless of its filename. MLS

The thing is that you are searching by name in the first place.

Don't think you really thought this through.

You are able to search by hash if you find the file on bitzi first and then press find more sources

[mls]

Originally Posted by Sleepless

The thing is that you are searching by name in the first place. Don't think you really thought this through.

Yes, I know you seach by name, and Limewire groups all files with the same name as one search result with # equal to the number of times that file name is found.

What I'm suggesting is that instead of grouping by filename, Limewire should group results by the hash code of the files found, so even if the filename differed it would still increase the # count.

Afterall, Limewire must know the hash code of each file it finds for a search in order to be able to right-click for Bitzi.

If I search for "Antmusic", I don't care if the file is called "Antmusic Track 4.mp3" or "Antmusic.mp3" or whatever, so long as it is the same MP3 file.

As I said, it's just a thought! MLS