SIR Users Conference
24^th September 2004
London

This paper will show how to use a set of members written by the author to access data via ODBC: it makes accessing ODBC sources as easy as accessing native SIR data. It will demonstrate setting up ODBC sources and retrieving data from them into a SIR program. It will also show the built-in ODBC importer for getting ODBC data into a SIR record.

Dave Doulton
University of Southampton

In the 1980s SIR/DBMS was available at data centres of several German universities. SIR was recommended to handle large and diverse datasets and to prepare those data for subsequent processing with statistical packages like SPSS. Compared to software like SPSS the usage of SIR itself was classified as 'complex'.

During the 1990s three things happened:

• The relational paradigm became predominant and products like Ingres and most of all Oracle spread widely.
• Microsoft Windows became a de facto standard. This led to a radical change in data storing and data usage.
• The marketers of statistical software gave the impression that complex data could be efficiently handled by the packages themselves.

Peter Lauf
Germany

I'd like to talk about the way I use SIR in my daily work, its indispensability in the process of manipulating data files, and the constructions and functions in PQL I like best. Emphasis will be on data manipulating, not data base design and data base administration. The talk is meant to be interesting to new SIR users in the first place; perhaps some old hands will be reminded of some useful PQL features they forgot all about.

Niels Veldhuijzen
CITO, Netherlands

SIR is being used for storage of data collected for the BCON Trial. The BCON Trial is a multicentre randomised phase III radiotherapy trial of locally advanced bladder cancer. The trial aims to address the problem of hypoxia by the addition of carbogen and nicotinamide to the radiotherapy. This paper will give some background to the trial, focusing on the SIR database structure and some aspects of data retrieval from a user perspective.

Kate Motohashi
Gray Cancer Institute

The maintenance of code has long been recognised as a serious problem in the computing industry. Trying to maintain large, badly documented systems which require substantial annual revision can be a nightmare.

SIR PQL has many very useful features which can aid the developer. I will show how the use of PROCEDURE, CRWARN, GLOBAL, CIF, DO REPEAT etc can all assist in producing more maintainable code. I will also show how it wasn't just the 'Year 2000' problem which can produce 'hidden time-bombs' in code and what can be done to minimise the problem.

Fran Williams
University of Essex

When SIR/DBMS became available on a variety of mainframe platforms in the early eighties, it was accessible in about half of the German university computing centres. Withdrawal of early support by social scientists, progress in computer science, strict application of German norms (such as DIN 2137), the rise of Personal Computing, and the success of graphical user interfaces ('point and click') and some other changes in the computer environment have made SIR databanks mostly 'legacy systems'. Consequently, SIR is little known and hardly used in Germany at present.

The possibility of a recovery of SIR in Germany, given the current administrative and legal needs, will be discussed from the perspective of a long-time, low-frequency user.

Klaus R. Allerbeck
JW Goethe Uni, Frankfurt, Germany