Sign up for our newsletter
Technology / AI and automation


Sometimes, jumping on to a marketing bandwagon can actually be detrimental to a company, and it is possible that Reasoning Inc’s ‘re-birth’ last year as a Year 2000 problem fixer, while giving it some marketing focus, may have actually deflected attention from the company’s real skills and goals as a code management and transformation company. The Mountain View, California company actually believes it has the tools to assist companies through any transformation they may embark on, mainframe to client-server, Unix to Windows NT, web-based transformations and any business transformations the future may throw at them, including Year 2000 and EMU Economic and Monetary Union system changes. Reasoning has been around since 1984, when it was set up by Stanford University graduates to build technology to enable the transformation of programming code from one language to another, for example from legacy code to Unix. It was funded initially by research grants and then by selling product – it actually sold to IBM Corp, which used the product to transform proprietary test scripts into AS/400 programs.

By Joanne Wallen

By last year it was in a make or break situation, and decided to go out and market itself in a much bigger way. It attracted Jim Treybig, founder and former chairman of Tandem Computers Inc, as its chairman, and he in turn went out and secured some venture funding for Reasoning; $2.5m from Informix Software Inc, Aspen Ventures and Novus Ventures (CI No 3,086). He also bought in a new senior management team, including Dr Tim Chou, formerly vice president of the server products group at Oracle Corp, as chief operating officer. Given Reasoning’s technology caters for code transformation, it realized that Year 2000 was a logical application for the technology and Chou admits it hung its marketing banner on the Year 2000 bandwagon. However, he is keen to point out that the company has much more than Year 2000 in its sights. The Code-Based Management System or CBMS – it’s supposed to parallel a DBMS or database management system only for code instead of data – enables legacy code to be drawn down into a central repository via a language gateway. Reasoning has written its own gateways for MVS Cobol, Ada and C, and partners have written gateways for PL1, Natural, Ingres ABS and Honeywell Bull Cobol. Once the code is in the repository, it can be queried and manipulated using Reasoning’s CQML, or Code Query and Manipulation Language.

Transformation Plug-Ins

White papers from our partners

A set of applications called Transformation Plug-Ins or TPI’s then effect the transformation of the code from one language to another, or from non-year 2000-compliant, to compliant code. New TPIs can be written, to support more programming languages, using Reasoning’s Software Developer Kit. Since the process is largely automated, Reasoning believes it has a pretty compelling story to sell to those businesses that have not yet started sorting out their Year 2000 problems, which according to Chou is a vast majority. Given the well documented shortage of qualified personnel available to fix the problem, Reasoning believes its automated code change will be a huge selling point. It estimates the number of lines of code a qualified programmer can fix is in the region of 100,000 to 200,000 lines of code per year. Reasoning’s software, it claims, will enable a programmer to get through 100,000 lines per week, and it has even achieved as much as 100,000 lines of code in three hours. The company has established Code Transformation centers in the US, currently one in California and one in Boston, and is about to set one up in Europe. These centers, it says, need only one or two code experts, and then a group of people who may even be fairly non-technical, but who can be taught to process the work the Reasoning software throws out. The company believes its software is extremely accurate at finding and fixing Year 2000 date problems for example. The software, which runs on RISC-based machines primarily under Sun Microsystems Inc’s Solaris, has an arithmetic confidence factor built in, so that if it does a pass through the data and is not 100% certain it has found a date field, it will flag these instances and pass them on for human intervention. Chou says some systems integrators, the company’s main target customers, are using the centers to set themselves up initially, and then they take over the work in-house. Others are outsourcing the whole transformation job. Earlier this year, Computergram questioned whatever will it [Reasoning] do after the Millennium? (CI No 3,086). Reasoning has no doubts. After Year 2000 and EMU, it will be busy with many other business transformations and legacy updates – with the pace of change in this industry, Java could be legacy by this time – and then comes the real future vision.

Central ‘code base’

Chou believes the next wave in business computing will see people realizing the value in their programming code in the same way as they see value in their data at present. Hence the DBMS analogies. At some point in the future, Chou says, if the code is in our CBMS, or central ‘code base’, why not leave it there? Then you can query it, catalogue it, re-use it, and generally treat it in the same way you treat data. One day, he says, you will see code as an item on the balance sheet. He goes further. The company’s ultimate aim, he says, is to become to code-base management systems what Oracle is to DBMS, a pretty bold claim for a company with just 90 employees worldwide whose immediate target is to launch into Europe and hit $7m revenue this year. However, long before this future vision can be put to the test, Reasoning is satisfied it will have its work well and truly cut out until the turn of the century, with ensuring that businesses code survives past midnight on December 31 1999.

This article is from the CBROnline archive: some formatting and images may not be present.

CBR Staff Writer

CBR Online legacy content.