[netbsd8]

Can anyone tell me why the designer calculate the credit modifier like this:

credit modifier = min{total uploaded *2 /total downloaded , squt(total uploaded +2)};

Is it optimal?

Thanks!

[qm2003]

[panicattack]
Yes, the modifier is optimal and there is no need to discuss the credit system any further.
[/panicattack]

[Nissenice]

^^

You CHICKEN, qm2003?

[netbsd8]

Why is it optimal?

For example: why not using *3 instead of *2? or why not choosing max instead of choosing min? or why limiting the upper bound of modifier to 10 not to 5 or 20......

Could you please give more detail or some related links?

Thanks!

[Ejack79]

Quote

credit modifier = min{total uploaded *2 /total downloaded , squt(total uploaded +2)};

The former presents the RATIO of the client's upload to his download,
the latter presents the AMOUNT of his upload to u.

For an example:
A upload 10M to u and download 5M from u;
B upload 100M to u and download 50M from u;
C upload 100M to u and download 100M from u;
then u can sort them as u like...

[raji]

One of the main issues in peer-to-peer file-sharing services is to ensure balanced transfers of data among peers. For instance, the popular eMule file-sharing client interface implements techniques based on a credit system and time modifiers, in order to guarantee a fair degree of contribution/reward, e.g., to prevent leeching. However, fixed heuristic modifiers may not be able to guarantee a fair treatment for every peer participating in the service. This paper introduces an optimized strategy to compute modifiers in order to promote cooperation among peers, while guaranteeing a suitable degree of fairness. Simulations are provided to compare the proposed mechanism to the standard one adopted by the popular eMule client interface.

This post has been edited by torpon: 11 November 2009 - 06:33 PM
Reason for edit:: No spam here, pleasw