bug-gne
[Top][All Lists]
Advanced

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

[Bug-gnupedia]So, how do we get started ?


From: Callum McKenzie
Subject: [Bug-gnupedia]So, how do we get started ?
Date: Wed, 17 Jan 2001 20:34:11 +1300

Since they mailing list archive seems to be empty, I'll write something
that addresses fairly basic issues and then get flamed because the real
discussions have already been going on for weeks.

OK, so we have some ideals, some very good ideals. But unless I'm missing
something we don't have any actual entries or a way of accessing them.

Content is easy (or at least getting a small collection to test the system
is easy), but since it's two in the morning the answer for me will be "not
right now".

The other side of an encyclopedia is its search function (even if it is
only the fact its sorted alphabetically). Ultimately we'd want a totally
decentralised peer-to-peer buzzword-of-the-hour application (both as a
standalone application and a "web-application"). But a first step will
obviously be to hand compile a list and post it to gnu.org. 

This immediately breaks the decentralisation rule until such time as its
mirrored. 

The second logical step is to a) create a script to update the list and b)
create a way for mirrors to regularly update their lists. Both these
should be easy.

After that you have to start thinking about real decentralisation of the
directory, how you're going to ensure data is mirrored (without
centralised control), blah, blah, blah, but thats beyond the
boot-strapping stage.

Also, you have to consider how to evaluate how accurate the information
is. You can come up with all sorts of amusing ratings schemes which will
be abused to hell. I have no good ideas yet, but it might be an idea for
authors who feel they may have a reputation concerning a particular
subject to gpg sign their contributions, so people can associate someones
reputation with the information they receive. Of course this limits
alterations (they would have to be done by anotation assuming the original
author is uncooperative). There are big flaws in this idea, but it doesn't
require any explicit infrastructure and it can't harm to play with it.

Anyway, I've already mentioned the time and so I'm off the bed, leaving
you to consider these deranged thoughts.

 - Callum





reply via email to

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