[Top][All Lists]

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

[GNUnet-developers] Re: finding files & database management

From: Igor Wronsky
Subject: [GNUnet-developers] Re: finding files & database management
Date: Sat, 13 Mar 2004 18:24:20 +0200 (EET)

On Sat, 13 Mar 2004, Krista Bennett wrote:

> > I don't really care too much about plausible deniability. I encrypt my
> > harddrive. With that out of the way, records of what I've indexed and how
> > would be really useful to me. I don't really care about records of what I've
> > inserted since, if I inserted it instead of indexing it, I may actually be
> > going for that plausible deniability thingy.
> Well, it's certainly something that could be done, and I'll see about
> putting something like that in this weekend if I get to it as an option
> for anyone who'd want to use it.

Let me bump in with a trivial but far-fetched fantasy.
The thing that would solve all these problems would
be a <drumroll...> content management wizzard! I have
personally been too lazy to do much of anything to
that end, my apologies, but I have a bit of a plan of
how it could look like.

The wizzard (lets have these double z's so as not to
confuse with windows wizards) would be a GUI app
(for example, extension to gnunet-gtk), that would basically
maintain a bunch of tree widgets and store them in
a $HOME/.wizzard/ directory on disk. Each tree would
either represent a namespace or a directory hierarchy.
For each tree node, the wizzard would know (perhaps
by asking from gnunetd) what is the nodes status (inserted,
indexed,lost,priority, and if indexed, is the file still
available locally, etc) and allow user to change these,
and change descriptions and keywords. Similarly new nodes,
subtrees etc could be added/removed to/from the trees and on
clicking 'commit' the wizzard would encode and send the
required parts of the updated structure to gnunetd (i.e., fresh
root blocks and the updated dir tree, not necessarily the
actual files).

Of course outside a regularly updating namespace, changing
the descriptions and keywords and/or file content would
cause keyword pollution, but thats something we can't
avoid in GNUnet. The only hope currently is that people
would start using namespaces and providing cross-links
to other namespaces with related interests. A thing like
content insertion wizzard might help the transition a lot.
If the wizzard were augmented with e.g. a feature that would
allow the user to maintain a 'friend list' in the namespace
(for example as a regular, standardized file entry in the
root-level directory, that would contain pointers to
other namespaces), this would work towards establishing
a link structure. gnunet-gtk could then try to fetch
these lists automatically or on click, allowing surfing.
I mean, perhaps in future, users are interested in
locating very specific content by keyword search.
But in the current phase I rather suspect that many
would find it acceptable if they could find something that
matches the overall 'genre' and be able to download the
results. Building a informal trust hierarchy between
namespaces is currently - to the best of my knowledge -
the best answer to content availability problem. By
this I simple mean that if person A finds a reliable
namespace B to his liking, and B points to some other
namespaces, it could be used as indication that the
other namespaces would have some reliability as well.

But I'd be quite happy with a realization of some
less grand scheme first. We don't necessarily need
to have the whole wizzardly corpulence to start with.

;) ;)


reply via email to

[Prev in Thread] Current Thread [Next in Thread]