Good programming practice and software selection

Approach for good programming practice

1. Plan before coding: understand the requirements and make a design (flowchart, pseudocode, state diagram) first.
2. Modular, top-down design: divide the program into small functions/modules, each doing one job (high cohesion, low coupling), with clear interfaces.
3. Structured programming: use sequence, selection and iteration; avoid goto and global variables; one entry and one exit per function.
4. Meaningful names and consistent style: descriptive names for variables and functions, consistent indentation and a coding standard.
5. Comments and documentation: a header per module (purpose, inputs, outputs, author, version) and comments explaining why, not what.
6. Defensive programming: check inputs and return values, handle errors and boundary cases, initialise variables, use watchdog timers in embedded code.
7. Reuse and portability: use tested library routines; keep hardware-dependent code separate; use constants instead of "magic numbers".
8. Reviews and testing: code inspection/walkthroughs, unit testing, and version control (Git).
9. Keep it simple; optimise only after measurement.

Criteria for selecting a software supplier (company)

1. Reputation and track record: years in business, references, number of similar installations, customer feedback.
2. Quality of the product: the company follows recognised process standards (ISO 9001, CMMI), has a testing and bug-tracking procedure, and published reliability or defect data.
3. Support and maintenance: helpdesk, updates and patches, response time, warranty, and long-term support.
4. Documentation and training: manuals, user training and source-code access/escrow.
5. Financial stability of the company, so that it will exist to support the product.
6. Compatibility and security: works with the existing hardware, OS and standards; security practices and licence terms.
7. Cost: total cost of ownership (licence, installation, training, upgrades), not just the price.
8. Trial/evaluation: demonstration or trial version and an acceptance test.