OracleFree calculatorPDF, Excel, Word exportOracle JD Edwards World

JD Edwards World Custom Object Conversion Effort Calculator

Custom code is usually the largest unknown in any JD Edwards World modernization. This calculator gives a first-pass estimate from object counts so you can see how much the decision to retire unused objects changes the effort.

Your numbers

programs
hours

Include analysis, rebuild, unit test and documentation.

objects
hours
screens
hours
interfaces
hours
30 %

Unused or replaced by standard functionality. Interfaces are not retired in this model.

0.5
$/hour
people

Your results

Total effort
5,364
Estimated cost
$455,940
Share carried forward
0.7
Build hours
3,576
Test hours
1,788
Elapsed time (140 productive hours per person per month)
9.6 months

Hours per object are planning assumptions. Replace them with measured figures from a sample of your own objects.

Save results:

Total hours vs share retired

Sweeps Share of programs, reports and screens retired as obsolete from 0 to 90, holding your other inputs fixed.

Custom code usage scan

Optional: a read-only scan of your IBM i custom libraries ranked by actual usage.

Optional. The tool and every export work without this. No spam.

How to use it

Count your custom objects: RPG programs in your custom libraries, DREAM Writer versions that are actually scheduled, custom display file programs and interfaces to other systems. Use WRKOBJ, DSPPGMREF and your job scheduler history to separate used from unused. Then enter hours per object type from a sample you have really analysed, not a hope. The retire slider models the share you decide not to carry forward.

The method

Build hours equal the carried-forward share of programs, reports and screens multiplied by their unit hours, plus all interfaces at their unit hours. Test hours are a ratio of build hours. Cost is total hours times rate, and elapsed months divide hours by team size times 140 productive hours per month. The chart shows how steeply total effort falls as you retire more objects.

  • Interfaces are kept at full effort because they carry data other systems depend on
  • Test ratio of 0.5 is typical for standard logic, higher for financial postings
  • Elapsed time does not include waiting for business users to test

Reading the result

If most of the cost sits in RPG programs, an object-by-object review for fit to standard functionality is the cheapest saving you can make. If interfaces dominate, the integration approach (APIs, files, middleware) matters more than coding hours. Use the three scenarios of 0, 30 and 60 percent retired to give sponsors a range, not a single number.

Common mistakes

Teams count source members instead of distinct business functions, forget that one DREAM Writer program may have a dozen versions in use, and assume all custom code must be rebuilt. Another mistake is applying the same hours to a 200-line report and a 5,000-line pricing program. Where you can, split your inventory in the companion worksheet before using the totals here.

Frequently Asked Questions

Does this assume a move to EnterpriseOne?

No. The hours are for rebuilding or converting custom functions on any target platform. Replace the unit hours with those for your chosen approach, whether that is E1, another ERP or a rewritten IBM i application.

How do I know which objects are unused?

Check job scheduler history, program call statistics where you have them, and ask the business. Objects not run or called for twelve months, including a full year-end, are candidates for retirement.

Why are interfaces not reduced?

Interfaces move data to and from other systems and are rarely safe to drop. If a partner system is being retired too, set the interface count lower.

Is 140 productive hours realistic?

It allows for meetings, leave and support work in a typical month. Use a lower number if your developers also carry production support duties.

If you want help producing an object inventory from your IBM i libraries and job history, Netray can run a read-only scan and give you a usage-ranked list.