Resembling problem are the most frequently approached case. What if both things are similar each other but little different? I like to write script code which fulfills the task automatically but sometimes it can’t. Once I’d got a same code block spreaded in various files and should modify something in the block, I integrated the block in a file and made refered files to include it. As you know, there’re various solutions case by case. The duplication problem can be divided into two categories: task dup and data dup. And here are the solutions I like the most.

Task dup
Task dup comes from misunderstanding or misleading. The case that someone misunderstands the requirement in accurate is a good example. I suggest three-step method to deal with this kind of duplication.

1. How to avoid? Should totally understand the problem, that is, should know the problem correctly.
2. How to deal with it? If the amount is big, try to automate to do such a task.
3. How to crack it? Record the reason and educate employees about the accident.

Data dup
Most data dups are trivial troubles. Script will help you. If paper record instead of electronic data is on the rise, no automatical way available. I can only show you an example of script-solved problem.

Example Case) A normal salary man John is working as an accountant. His daily work is to record finance many clients defray. One day, around the time to leave the office, he noticed that he’d got a mistake of inputting the client code incorrectly (he added 1 to all of the codes). So he asked his co-worker Ted, a good scripter, to modify mis-coded data automatically. Ted simply made a script for him as below.

def convert_client_code( miscode, other_args ):
if miscode > 0:
miscode = miscode + 1
# record the code into a file in a format

There’re so many tools or methodologies to manage project. But they are not sufficient to judge whether the project goes well. Totally integrated data collection, measurement, analysis, and process would make it better. First of all, monitoring the trend of customers and interpreting it are the preceding task. (But never reveal them to the public as your data collection should keep their trust) That analysis can help estimate the project objectively trusted. Second, communication matters on the project while business is separated development or product line. (or vendor to customers) If one of them are bottle neck, then whole project would be delayed. Thus you’d rather share all the things of the project in your company so that project members can catch the narrow position. Lastly, you should avoid the following situations:

• Ambiguous or diffuse project ownership.
• Stakeholder complexity
• Conflicts of interest
• Inconsistencies (hardest to discern in advance)

Project management is not the easy one and the failed project is desperate. The best is not to start the project if it has less possibility or is in a bad condition.

Requirement is a kind of communication. Properly documented requirements help reduce the gap between stockholders (who mostly require). And understanding requirements from the start synchronize development group with business users and their needs. The following is the list of proper solutions.

• Document the requirements clearly.
• Set priority of each requirements.
• Record the trace of the requirements.
• Technical team should communicate to business users.
• If there lies any change, be clear about the implications for scope and cost.

These rules are not the essential conditions to get the project succeed. But they’re some of guidelines which can avoid project getting down into failure.