Re: [dijjer-user] Dijjer more efficient than Bittorent?

[Ian C. <ia...@lo...>]

On 16 May 2006, at 16:45, chris wrote:
> Ian Clarke wrote:
>> On 16 May 2006, at 16:13, chris wrote:
>>> Ian Clarke wrote:
>>>> <snip>
>>>> Files initially come from the web server, but subsequently they   
>>>> come  from peers in the network that are caching pieces of the   
>>>> file.  For a  file of 1GB, I would say that the web server  
>>>> hosting  the file might  be expected to have to send 2-3GB of  
>>>> data in  total, even if there are  thousands of downloads.
>>>>
>>> How is this more efficient than  Bittorrent?
>>
>> I already tried to answer that question.  If my answer was  
>> inadequate  then please explain why, and I will try to clarify.
>
> Thank you for your patience. I think I may not have worded the  
> question properly, so here goes, I understand that dijjer spreads  
> the load differently than Bittorrent but to take your example does  
> this mean that for a file of 1GB would the hosting web server be  
> expected to send more than 3GB when using bittorrent as opposed to  
> less than 3GB when using dijjer.

The simple answer is that I'm not sure whether Dijjer or Bittorrent  
will put more or less load on a central server when all other factors  
are equal, the only way to know this would probably be to do some  
testing under controlled conditions.  I think in both cases, however,  
the amount of load should be negligible (3X the size of the file  
should not be a lot of bandwidth proportionally).

Ian.