Re: [unrev-II] DKR for Open Source: Viability

From: Jeff Miller (
Date: Mon Feb 07 2000 - 18:59:26 PST

From: Jeff Miller <>

On Tue, Feb 08, 2000 at 12:02:12PM +1100, David Robinson wrote:
> This brings up the question:
> How do we want to store the information?

> Things get very complicated when everything is being stored in one BIG
> If things can be distributed over many databases in many locations but
> all interrelated one failure doesn't mean everyone looses out. Each
> module or NIC could be a database.

A distributed system such as the web is now. With replication (read
smarter mirroring) so when one site goes down the secodaries are
automatically tried. Our not only needs to be forward and back, but across
such that when the document is updated on the primary or mirror sites then
the others get updated as well. This raises questions of versioning and
resolving discrepences.

so linking may go like this

pribase =
secbase =
secbase =

<backlink src=anotherdoc/page.dkr#para>

<forlink src=fredsdocs/testing.dkr#here>

> We should use a TOP level Directory Like NDS (Novells Directory
> Database) which allows all databases to know the basics about all other
> databases so when information needs to be gathered or stored it can use
> this and contact the correct server quickly.

I'm not familiar with NDS, but from what I hear it's similar to LDAP. But,
is this really nessacary.

> We can create a hugh information repository but keep it manageable as
> each NIC/database/module/group knows the ins & outs of their system. But
> may not know how the whole system works and doesn't really need to.


> Unlike the web is today where it can take hours to hunt down the correct
> information, when a user accesses the database and enters a search query
> for that database/NIC our system can check the Top Level Directory and
> suggest other areas that the informa tion might be stored.

I don't think will every solve the problem of spending hours to look for
the correct information for several reasons including commerial interests
bending things to get more people to see things their way and people think
in different ways about the same subject using different words for the
same thing making hard to search for that subject in a sea of information
about similar or different things with similar words.

> I believe we have to write our own Interface that perfectly suits what
> we want to do.
> When I'm creating references to other documents I want to just drag that
> document in or enter its ID tag or category tag so others can easily go
> off and find out why I came to that conclusion. I don't want to type out
> 10 URLs which in 6 months time my n ot exist.

I like this idea. I hate entering references into documents because it
tedious, error prone, and who ever you present the finished article has
their own views on how thing should be referenced.


--------------------------- ONElist Sponsor ----------------------------

Shop for your Valentine at eGroups.
<a href=" ">Click Here</a>


Community email addresses:
  Post message:
  List owner:

Shortcut URL to this page:

This archive was generated by hypermail 2.0.0 : Tue Aug 21 2001 - 18:56:44 PDT