DataMuseum.dk

top - metrics - download

    Length: 6813 (0x1a9d)
    Types: TextFile
    Names: »q-introduction.tex«

└─⟦3d0c2be1b⟧ Bits:30001254 ISODE-5.0 Tape
    └─⟦eba4602b1⟧ »./isode-5.0.tar.Z« 
        └─⟦d3ac74d73⟧ 
            └─⟦this⟧ »isode-5.0/doc/manual/q-introduction.tex« 

TextFile

% run this through LaTeX with the appropriate wrapper

\f

\chapter {Overview of QUIPU}

\f

\section {Summary}

QUIPU is a public domain implementation of the OSI Directory
as defined in \cite{CCITT.Directory,ISO.Directory}.
The operations of the QUIPU Directory are aligned to that of
\cite{CCITT.Directory} which is the most recent text, this is expected
to be the same as the forthcoming IS which is currently available in its draft
form \cite{ISO.Directory}.

It is intended to provide an environment
for early experimentation with standardized Directory services.
It is used by the ISODE for identification of the location of OSI
applications (including QUIPU) and for provision of white and yellow
page services.
The Directory Abstract Service and DSA Abstract Service
defined in \cite{CCITT.Directory,ISO.Directory} and their associated 
protocols are supported.


Major aspects of the QUIPU implementation are:
\begin {itemize}
\item
Use of  memory structures to provide fast access
\item
Activity scheduling within the DSA to allow for multiple accesses
\item
General and flexible searching capabilities
\item
Extensions to provide access control
\item
External schema management
\item
Use of the Directory to control Distributed Operations
\end {itemize}

The current implementation provides a DSA, and a procedural interface to the
Directory Abstract Service, which will enable other applications to use the
Directory.
There is also a DIrectory SHell interface --- DISH.  This provides full
access to the Directory Abstract Service, using the procedural interface.
Standard Distributed Operations are used with both referrals and chaining
(using the Directory System Protocol) provided.

\ifnum\volnum=5
A full discussion of the design issues relating to QUIPU can be
found in \cite{QUIPU.Design}.
\fi

\f

\section {Pronouncing QUIPU}

The name of the INCA Directory is QUIPU.
The official pronunciation of QUIPU takes two syllables:
{\em kwip-ooo}.

\f

\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 {\em Quipus}.
The encoding was obscure, and could only be read by selected trained people:
the {\em Quipucamayocs}.
The Quipu was a key component of Inca society, as it contained information
about property and locations throughout the extensive Inca empire.


\f

\section {Objectives}

\subsection {General Aims}

QUIPU has a number of general aims:

\begin {itemize}
\item To produce an implementation which follows the
emerging OSI Directory standards.  

\item Flexibility to enable the system to be used
for experimentation and research into problems relating to Directory Service.

\item Investigation of distribution and replication

\item Pilot experimental usage.
\end {itemize}

\subsection {Technical Goals}

The major goals of the QUIPU Directory Service are:

\begin {itemize}
\item
Full support of the Directory Access Protocol,
Directory System Protocol and Distributed Operations, as 
defined in \cite{CCITT.Directory}.
\item
Support of the majority of the service elements specified in
\cite{CCITT.Directory}.
\item
Ability for interworking with other Directory implementations, including
use of
referrals and chaining.
\item
Very full searching and matching capabilities, beyond the minimum
required by \cite{CCITT.Directory}.
\end {itemize}

The following are not goals:

\begin {itemize}
\item
In practice, the memory based approach has led to a quite fast lookup and
searching.
\item
The ability to handle very large volumes of data (e.g., greater than 100~MB
or 1 Million entries per DSA) is not a requirement.
\item
Substantial data robustness is not required: there is no need to employ
complex data backup techniques.
\item
Use (as opposed to provision) of Authentication services.

\end {itemize}

\f

\section {QUIPU Support Address}\label{quipu:support}

If you have any problem installing QUIPU,
following the documentation
or any other QUIPU related problems, then there are two 
discussion lists.

Comments concerning the operation of 
QUIPU should be addressed to the QUIPU support address:
\[\begin{tabular}{ll}
Internet Mailbox:&      \tt quipu-support@cs.ucl.ac.uk \\
Janet Mailbox:&    \tt quipu-support@uk.ac.ucl.cs \\
X.400 Mailbox:&	   \tt surname = quipu-support \\
& \tt ou = cs \\
& \tt Org = UCL \\
& \tt PRMD = UK.AC \\
& \tt ADMD = Gold 400 \\
& \tt C = GB
\end{tabular}\]

There is also a discussion list for a general discussion of topics
related to QUIPU; the address is as above, but with ``quipu-support''
replaced by just ``quipu''.
We suggest that everybody who is intending to run QUIPU should be on this
list, as this will be used to keep you informed of what is happening.
Details of updates will also be sent to this list.

To be added to the \verb"quipu" discussion list, send a message to 
``quipu-request''.

\f

\section {Acknowledgements}

QUIPU was developed at the Department of Computer Science at University
College London, under the {\ae}gis of the INCA\index{INCA} (Integrated Network
Communication Architecture) project, which is project~395
of ESPRIT\index{ESPRIT} (European Strategic Programme for Research into
Information Technology).   The partners of INCA (GEC plc\index{GEC plc},
Olivetti\index{Olivetti}, Nixdorf AG\index{Nixdorf AG},
and Modcomp GmbH\index{Modcomp GmbH}) are acknowledged for releasing this
software into the public domain.
The Joint Network Team (JNT)\index{JNT}, 
for continued funding of QUIPU.

Marshall Rose\index{Rose, Marshall T.} and 
Chris Moore\index{Moore, Christopher W.} of 
The Wollongong Group helped considerably in the development
of QUIPU, and integration with ISODE.
Simon Walton\index{Walton, Simon} of University College London,
also provided much help in integrating the software with ISODE.

Steve Titcombe\index{Titcombe, Steve}, a student at 
University College London, did much 
of the early work on DISH.
Andrew Eliasz\index{Eliasz, Andrew} of University College London developed the SUNINT interface.
Paul Sharpe\index{Sharpe, Paul}, of GEC Hirst Research Laboratories
put considerable effort into the development of WIDGET.

Mike Roe, \index{Roe, Mike} Stella Page \index{Page, Stella} and Alastair
Hickling \index{Hickling, Alastair} of University College London
helped considerably with testing the software and proof reading this manual.
George Michaelson\index{Michaelson, George} 
of the Department of Computer Science at the University of Melbourne in
Australia and Julian P.~Onions\index{Onions, Julian},
at the Department of Computer Science at Nottingham University, both ran
early versions of the system, and provided much useful feedback.