There are a few main reasons why legacy code modernizations fail.
First, manual re-writes can take a great deal of time, particularly for large applications. During this time, changes to the project goals may cause further delays or even cause the project to restart from the beginning. The cost for manual rewriting of code is already very high, and so starting over becomes particularly costly.
Also, long code freezes of months and years are common with manual or semi-automated conversions, which result in major issues for your user base, and eventually project failures.
Transliterative tools that focus on line-by-line translations of the source code into the modern code are also widely used. Just as using a dictionary to translate from another language to English one word or sentence at a time would result in many areas of confusion, so these tools produce results that don’t accurately reproduce the functionality of the original application. Even when completed, this results in complicated and unmaintainable applications.
Finally, the costs associated with large numbers of people on partially-automated and manual modernization projects can be very high. These costs grow as features are added or changed several months or years into the project, until millions and often hundreds of millions have been invested.
The fully 100% automated functionality of TSRI’s JANUS Studio solves all of these problems, and is the reason why all of our past projects were successfully completed, and all of our past projects are reference able. This perfect track record of success has been honored by awards ranging from the Stevens award for our CEO Philip Newcomb, as well as the Top 20 Most Promising Defense Technology award from CIO Review.