[E-Lang] A more critical question (was: an example impatience policy )

[Tyler Close]

Zooko wrote:
>, and a
> working "introducer" (which is called a "MetaTracker")
> which maps public key
> ids to comms addresses and a few other bits of generally
> useful info.

This sounds like the VLS. I assume that the MetaTracker is what is
documented at:

http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/~checkout~/evil/hackerdo
cs/distributed_metatracking.html?cvsroot=mojonation

If so, then the "white pages" part of the MetaTracker is the VLS
functionality. I think the "yellow pages" part of MetaTracker is best
implemented in an application that gets layered on top of the base VLS
service.

Unfortunately, it looks like the erights.org site doesn't have
anything at:

http://www.erights.org/elib/distrib/vattp/VLS.html

My knowledge of VLS was obtained directly from the horse's mouth.
Armed with this knowledge, I attempted to generalize the VLS into a
service that could be used by all comm systems. My HTTPY design was
developed on this list starting at:

http://www.eros-os.org/pipermail/e-lang/2000-May/003340.html

Based on my reading of the MetaTracker proposal, it is missing the key
feature that makes VLS so scaleable, embedding the list of TCP/IP
addresses of the VLS servers (MetaTrackers) to consult directly in the
target URI. Reading the httpy:// thread (link above) should give you
the rest of the context. I can probably fill in any holes.

> However what we, and everyone else, need is a distributed,
> attack-resistant,
> scalable mutable namespace, which seems like a difficult
> problem and perhaps
> not even feasible.

I think it is not only feasible, but has a quite simple and elegant
solution. The communities.com people really did a good job on the VLS
concept. I think a generalization of it could be the base building
block that we all need. It sounds like Mojo has yet to fully implement
a solution here, so now seems like a good time to merge designs.

> I'm actively talking to various people who are interested
> in that project.
> There are *many* people thinking about variations on that
> theme and many of
> them don't know about each other or about the shape of the
> design space that
> they are exploring.

Invite them in.

Tyler