ENCE ON ACHIEVING QUALITY IN SOFTWARE, It is not easy to get the proceedings of a conference published as a book nowadays, unless the publisher can be sure that the finished item will be included in delegates' conference packs, thus ensuring at least a hundred-odd sales. Certainly I received this book for review several weeks before the actual conference took place. In some cases there must be a danger of sacrificing quality in this scramble to publish: some papers will be rushed or will fail to materialize at all. In this volume, the invited paper of Marc Aye1 entitled 'Verification and validation of knowledge-based systems' is a mere 11-line abstract and Brice Lepape's invited paper is a bland, two page overview of the ESPRIT programme, far less informative than the official web site. However, apart from this and the inevitable inconsistencies of camera-ready publication, this book has much in its favour: the 29 non-invited papers are usefully grouped into sections (process improvement, quality practices, software testing, numerical assessment, quality modelling, object oriented, KBS quality, formal methods and two sections on quality measurement); there is a keyword index; the papers represent ten different countries, including three each from the U.S.A. and Japan, and offer a good balance of academic and indus- trial authors; the conference itself is the third of a reputable series which takes place biennially in Italy under the auspices of IFIP. Still, one might want more compelling reasons for buying this book, especially at €60, or for reviewing it in STVR for that matter; and in fact the reason I chose to review it in this double issue was that almost half the papers deal with testing, verification and reliability, or with object-orientation, or both. So I shall concentrate on these papers and then leave the reader to judge for him or herself.
Part four, entitled 'Software testing', has three papers which deal interestingly with three very different aspects of this subject. Mirella Mastretti