⟦7cfb8a793⟧

TextFile

\chapter {Overview}

\section {Introduction}

QUIPU is a Public Domain implementation of the OSI Directory as 
specified in CCITT X.500 Recommendations /
ISO DIS 9594 for Directory Services \cite{ISO.Directory}
\cite{CCITT.Directory}.
It is intended to provide an environment for experimentation and for early
pilots using Standardised Directory Services.  QUIPU is currently aligned to
the CCITT X.500 Recommendations (unofficial version).  This is expected to
be technically identical to the ISO IS.  The latest available ISO version is
the DIS.  

This document describes the design of QUIPU 5.0, and looks at some
extensions which are planned for QUIPU 6.0. 
The reader who is only interested in the first aspect, should be careful to
note those aspects (clearly marked) which are not yet implemented.
This document is intended to complement
the ISODE Manual \cite{QUIPU.Manual}, which describes how to operate QUIPU, and how
to interface applications to it.  
Familiarity with the OSI
Directory specifications is assumed \cite{CCITT.Directory}.


QUIPU fully implements both of the OSI Directory Protocols, and a number of
extensions.
The highlights of the QUIPU Directory Service
Implementation are:

\begin {itemize}
\item
Use of memory structures to provide fast access, without use of
complex keying techniques.
\item
Activity scheduling within the DSA to allow for multiple accesses.
\item
General and flexible searching capabilities.
\item
A mechanism to provide non-local access control.
\item
A mechanism to provide external schema management.
\item
A sophisticated approach for management of distributed operations and
replication.
\end {itemize}

The current implementation provides a DSA, and a procedural interface to the
Directory Abstract Service and the associated Directory Access Protocol
(DAP), which will enable other applications to use the Directory.

\section {General Aims}

To understand the rationale behind some of the decisions, it is
useful to consider the original aims of the QUIPU project.
These
can then be mapped onto a number
of more technical considerations:

\begin {itemize}
\item To produce an implementation which followed the
emerging standards.  This is an aim in itself.

\item Flexibility, to enable the system to be used
for experimentation and research into problems relating to directory
services.

\item To provide a vehicle for experimentation in the area of
distribution and replication.

\item To provide some level of real usage.
This sort of work is useless if entirely confined to the laboratory.
It is important that it is capable of use for some level of experimental
service.  However, it is not consciously designed to evolve into a full
fledged product.
\end {itemize}

As the work has evolved, the following goals have emerged as
additional to the original ones listed above:

\begin {itemize}
\item To provide a public domain the OSI Directory implementation as a part of
the ISODE package.

\item To provide integrated support for the ISODE Applications.

\item To be used as a part of the initial pilot Directory Service in
the UK Academic Community and in other pilots.
\end {itemize}


\section {Technical Goals}

The major goals of the QUIPU Directory Service are:

\begin {itemize}
\item Full support of the Directory Access Protocol and Directory System
Protocols \cite{CCITT.Directory}.
\item
Support of the majority of the service elements specified in the OSI Directory.
\item
Full interworking with other OSI Directory implementations.
\item
Very full searching and matching capabilities, beyond the minimum
required by the OSI Directory.
\item
Provision of a system which has potential for very high distribution.

\item Support of distributed operations in a manner which is full
conformant with respect to non-QUIPU systems, and provides additional
functionality for QUIPU systems.

\end {itemize}

The following areas were not intended as goals in the initial system.
Some discussion is given as to how these areas might be tackled in
future versions.

\begin {itemize}
\item
The QUIPU Directory is not intended for very large scale
systems (i.e., Millions and tens of Millions of entries per DSA or hundreds
of megabytes of data per DSA).

\item
Substantial data robustness is not required: there is no need to employ
complex data backup techniques, such as replicated hardware.
\item
The security aspects of the OSI Directory were initially omitted, as not
required by the general aims.   
At this point, there is no reason why this aspect should not be
integrated.

\end {itemize}

\section {Further QUIPU documents}


The following documents are available:

\begin {itemize}
\item This document, which describes the design of QUIPU \cite{QUIPU.Design}.

\item The QUIPU Manual, which describes how to use QUIPU \cite{QUIPU.Manual}.

\item A paper on the original design, which is mainly of historical interest
\cite{ECW87.INCA}.

\item A paper presented at the 1988 IFIP 6.5 conference, which gives a
general overview \cite{QUIPU.IFIP}.

\item A paper presented at Esprit Conference Week 1988, which describes the
distributed operations \cite{QUIPU.Distributed}.
\end {itemize}

All of these papers, except the third, are distributed online with QUIPU.

\section {Acknowledgements}

QUIPU was developed in the Department of Computer Science at University
College London, under the {\ae}gis of the INCA (Integrated Network
Communication Architecture) project, which was project~395 
of Esprit  (European Strategic Programme for Research into
Information Technology). 
The partners of INCA  (GEC plc, Olivetti,
Nixdorf AG, and Modcomp GmbH) are acknowledged for releasing the first
version of this
software into the Public Domain.

Continued funding of QUIPU as Openly Available Software is provided by the
Joint Network Team.

Colin Robbins, Alan Turland, Alastair Hickling, Marshall Rose, Stella Page,
and Mike Roe have all made useful comments on this document.   

QUIPU 5.0 was implemented
primarily by Colin Robbins and Alan Turland.  Additional acknowledgements
for implementation efforts 
are given in the manual \cite{QUIPU.Manual}.

\section {Pronouncing QUIPU}

It is clearly important to distinguish QUIPU verbally as well as in writing.
The official pronunciation of QUIPU is {\em kwip --- ooo}.

\section {Why QUIPU}

QUIPU was originally developed as a part of the INCA project.
The Inca of Peru did not have writing.  Instead, they stored information on
strings, carefully knotted in a specific manner and with coloured thread, and
attached to a larger rope.
These devices were known as ``Quipus''.
The encoding was obscure, and could only be read by selected trained people:
the Quipucamayocs.
The Quipu was a key component of Inca society, as it contained information
about property and locations throughout the extensive Inca empire.
DataMuseum.dk

DKUUG/EUUG Conference tapes

⟦7cfb8a793⟧ TextFile

Derivation

TextFile
