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

Re: [xanadu] flecks ?????



Andrew Pam wrote:
> 
> I'm disappointed in you, Ian!  As a long-time Xanadu follower, you should
> well know that the Xanadu design evolved beyond the idea of a proprietary
> network since at least 1993-94.

I'm also a long-time Xanadu follower and I was unaware of the shift in
direction.  I knew it was needed but haven't come across anything from
the Ted Nelson et al stating so with some authority.

The direction I'm taking the open Xanadu/Udanax sources is a
decentralized,
non-proprietary anonymous collection of backend servers that:

a) fully implements Ted's xanalogical storage system

b) gates in a limited fashion to the HTTP world, with links over text
   being translated to HTML on the fly, for those that are relevant.

c) swallows email/newsgroups in massive quantities

d) provides a variety of APIs for accessing the docuverse

e) layers onto freenet the BEBE protocol in Gold/email archives

f) comes with sample browsers in wxPython and GTK/C++.

Crypto IDs (ala 'brands' in _Earth Web_) will be provided by the browser
and be compatible with those in freenet.

One area I've wrestled with is the Xanadu idea of royalties.  A central
billing system is a threat to privacy.  Not keeping detailed records of
what someone reads improves their privacy but also offends some people
when the bill comes and they wish to know what they're paying for.  But
keeping itemized charges is offensive to others.  The only solution I
see is to a) make most documents available for free, like the current
web, and b) decentralize the billing by making it a crypto deal btw
the reader and the author, and let the reader's browser provide the
itemized charges *exclusively*.

I wish Ted Nelson and others in the ***HEART*** of the architecture
would join in a discussion of how Xanadu is evolving today.  Most of
those I talk with are either non-technical Xanadu groupies waiting
for the Second Coming or those who think XML/SQL/Java are _the_
solution (the Open Hypertext System vision at www.bootstrap.org).

<sigh>

In any case I am proceeding onward and making everything I do
available online at http://www.timecastle.net/v/xanatalk.  I would
welcome fellow code spelunkers/documenters, as that is the current
task.  And there are some *quite* interesting algorithms in Gold.
Most intriguing indeed... ;-)

> The present design has been based on a
> set of individually specified but integratable modules, some of which
> already have open-source prototype implementations, for some years now.

Besides the open-sourced Green and Gold implementation, made public just
one year and 25 days ago, what other implementations have been public
for "some years now"?

-Jeff Rush