I noticed a paper called THE COMING REVOLUTION IN LIBRARY SOFTWARE
by David Dorman on IndexData’s
Web site. I checkout this site regularly because they provide good, reliable, efficient open source technology for search and retrieval of meta-data with packages such as Yaz
. In the paper he argues a new business model for delivering software to libraries called commercial open-source will cause a paradigm shift in the market. He makes a plea for libraries to fund the initial development costs with a 10 point plan. He sees further development and support being charged to customers. This doesn’t seem so different from a traditional commercial model where development costs are recovered from customers through purchase costs and further development is funded through recurrent payments. He claims open source software development results in ‘less expensive and better designed software, and speedier development’ than development by traditional vendors. In the competitive market of library software supply this seems difficult to justify. Competition is driving down customer costs because cost is a factor in winning bids. ‘Better designed’ software can give more reliable and usable solutions generating fewer support calls. Lower maintenance costs are attractive to vendors because it reduces costs making lower recurrent charges possible. Lower total cost for customers over the lifetime of a system is an important competitive advantage. Vendors seek speedier development to reduce time to market as competitors fight to attract new customers. He says for commercial open source vendors ‘Development, rather than being an opportunity to sell more licenses, or a burdensome overhead cost to be avoided if possible, becomes the primary revenue generator’. My experience with Talis is we enjoy our development; our purpose is to develop solutions for our customers. And if we don’t develop attractive solutions we don’t sell so development is central to our success. On quality he implies peer review of open source code is more likely to achieve high quality software than a software engineering process incorporating reviews at each major milestone of analysis, design, implementation and test. In addition to software development processes Talis publishes end user documentation, database schema, stored procedure code, scripts for useful utilities. He suggests abandoning proprietary development tools in favour of open source alternatives. We use third party programming tools such as Microsoft’s Visual Studio, a best of breed tool, to speed delivery and enhance quality. On commercial open source he says ‘This Model requires a new and closer relationship between vendors and libraries’. Isn’t this what all vendors are striving for? Talis fosters a community who share useful tools in source code provided by customers and partners through our Talis Developer Network
. His vision assumes there are cohorts of willing library programmers with sufficient knowledge, skill, free time and resources to develop library software. Instead I see a community of customers behaving with enlightened self-interest to pass on experiences to fellow customers. In practice my guess is IndexData will have a core development team with a few external trusted developers providing code fixes. I don’t see a paradigm sift, I see another competitor in the library software market.