There is an expression in software development that goes like this:
There is never enough time to do the job right,
but there is always enough time to do it over!
We believe every job should be done right the first time.
We take pride in knowing that every detail is examined and programmed correctly.
We would rather deliver a working, bug free, program even if that
means it has fewer features, than one loaded with features that do not work.
The right way to build software:
Do you need an application written or advice on some aspect of computerization?
Please contact us with your requirements so that we can exhibit our philosophy in action.
We can also provide references on request.
- Discuss the requirements.
Seek the crux of the problem; the main reason the project is being requested.
Be sure to understand the real value of the application to the user.
Do not dwell on useless frills.
Do look for ways to add value that the user may not have thought of.
Avoid duplication of data entry and/or data storage.
- Design it all then build the essential.
Do a proof of concept prototype. Build the important function(s).
Be sure they are complete, reliable and fast. If the program is not responsive
(with real data), then the program is useless.
Design the user interface and discuss with users. Check that it is complete
and logical (from their perspective, not yours).
Make input easier by providing reasonable default values.
Suggest additional calculated information such as percentages, maximums and averages.
- Write the program and be prepared to change.
Discuss the program with the users at every stage.
The user often does not really understand what he is getting until he sees
it in operation.
Be prepared to modify based on their feedback. This may be true even after the
program is "complete".
- Be available for support.
Answer the telephone. Answer the e-mail. Get back to the caller/e-mailer.
If something is not right then fix it. Assigning blame is not the issue.
Whether there is an error in the program, the user made a mistake, or there is a
misunderstanding regarding the program's function, get to the bottom of the problem
and provide a solution.
Revised: 2001 Feb 18
Jay Schwartz, President
Elipse Home Page
Send us e-mail