[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ih] reinventing the wheel, was Internet History Lives on the Internet?

On 02/26/2019 03:08 PM, Jack Haverty wrote:
> If subsequently someone in EU attempts to download that same file, and 
> the first EU downloader is still online and has the file, the file will 
> hopefully be transferred from the first EU site to the second.

Hopefully is the operative word.  The first downloader in EU may be a 
leacher and / or otherwise prevent people from retrieving the file from 
them.  (Rate limiting to something ridiculously slow also counts.)

If the machine in CA has crashed and no longer able to serve the file, 
the 2nd downloader is in a bad position.

Assuming that there is nobody else (critical mass) with the file.

> In that scenario, there is no need to transfer the same data across the 
> US and Atlantic a second, or third, etc. time.

There may very well be.  Especially with leachers that don't play nicely 
with others.

> It's already in EU.  Those redundant 10,000 mile transfers are eliminated. 
> The more players, the better it presumably gets.

Hope for it.  But don't bet on it.

> I think that's what Miles means by "effectively multicast". It's the 
> characteristic that makes BT attractive for things like distributing 
> DVD-scale ISO files of Linux distributions.

Multicast has an inherent timing component to it.  So if the 2nd 
downloader isn't online to receive the multicast when it's sent, they 
wouldn't get anything.

BitTorrent is not multicast.  BitTorrent is not effectively multicast. 
BitTorrent is not remotely multicast.  (At least not by any definition 
that I'm aware of.)

BitTorrent is an efficient way for multiple people to efficiently share 
files.  But that is not sufficient to call it multicast.

Grant. . . .
unix || die