[mnet-devel] A brief bit of history... [was Re: ecash]

[Jim McCoy]

On Wednesday, February 19, 2003, at 10:20 PM, Artimage wrote:
[...]
> Besides the fact that they didn't solve a specific problem. (I'm not
> saying they didn't do cool things. But no one was hurting for a 
> solution
> to the problems they seemed to solve. Or, also possible, I don't
> understand the problem they were aiming at.)

Ding! Ding! Give that man a cigar...

There are really lots of problems to be dealt with in this space, 
ranging from legal/regulatory woes (are you a non-bank financial 
instrument or not, etc.) to the nitty-gritty implementation problems, 
to the inevitable "so what?" question to be answered.  While I will 
avoid speaking about the Digicash case [okay, I won't :) -- DC was too 
early and did not want to deal with porn clearing, which is really the 
only money game in town at the time] I can say that part of the MN 
problem was looking at "p2p" too early.  We were building a distributed 
storage architecture before Napster ever existed as the first step 
towards creating a sort of cypherpunk grid-services architecture to use 
a few buzzwords people might be more familiar with -- to put this in 
context, Doug and I were sketching out the basics of what would 
eventually become MojoNation back in '92 and were white-boarding ideas 
that later become MN in '95...

The storage back-end (what eventually was MN and now lives on in 
HiveCache and Mnet) was needed as the first step towards a truly secure 
anonymous email solution I was working on. ZKS had received funding and 
seemed to be making waves, so we thought that a truly secure version of 
a penet-style anonymous email system might be a good idea.  Besides, 
distributed storage had lots of potential uses if the mail idea didn't 
work.  To do a decent distributed storage system among a group of 
completely untrusted peers you need some sort of distributed resource 
allocation mechanism, thus the need for Mojo.

Along the way Napster rose up and we were swept along in the wake, 
thinking that Mojo would solve what was perceived as a big problem at 
the time: too many downloaders and not enough servers.  Part of what 
made Napster possible (and why it didn't happen before, even though the 
ideas behind it are older than dirt and its architecture was only a 
small step beyond the simplest possible solution to the problem) was 
the fact that people suddenly had access to MP3 encoders and could rip 
CD collections.  When Napster first appeared these encoders were 
complicated command-line tools so most people did not have their own CD 
collection ripped yet, with Napster's arrival it was easier to download 
your collection than to rip your own CDs. The consequence was that lots 
of people using Napster ran into slow downloads, overloaded servers, 
and a generally unsatisfying experience.  Our new thinking at the time 
was that Mojo could solve this by compensating those providing the 
resources (in this case, bandwidth and the time needed to encode a 
music collection). By the time we had everything ready this problem had 
pretty much become a thing of the past. Encoders were a part of every 
music app and most people had already built up a sizable MP3 collection 
from either Napster or their own CDs.

The first question should always be: what problem are you trying to 
solve here.  This is why I think that micropayments and other 
distributed resource allocation systems are not a critical need for 
Mnet, but they might be worthwhile in the future and it is probably 
easier to re-attach the old code and keep it up to date than to rebuild 
this particular wheel when the time comes.  This is also why I think 
the paris metro pricing model and "pay for preference" rather than "pay 
to play" that MN had at the end is going to be more useful and relevant 
to this discussion than the traditional ecash fantasies that sometime 
pop up when talking about this subject.

Jim



-------------------------------------------------------
This SF.net email is sponsored by: SlickEdit Inc. Develop an edge.
The most comprehensive and flexible code editor you can use.
Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial.
www.slickedit.com/sourceforge
_______________________________________________
mnet-devel mailing list
mnet-devel at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mnet-devel