Over the years I have observed and experienced many different ways to select technology and architect solutions. I thought I would share some of them today. Please note, these are not necessarily recommendations; just observations of real world decision making.
- Follow the vendor. This methodology is based on the vendor who provides the most fringe benefits to the decision maker or in in “free” technology enhancements. Free golf, drinks, dining, sporting events, fancy hotels are often contributing factors to technical decisions. Others throw in additional packages, hardware, or consulting services to help you get your solution on the road. I have experienced this from third party software and hardware vendors, and directly from software platform manufacturers.
- Follow the expert. This methodology follows the recommendation of experts in the form of technical articles, presentations at conferences, books, training courses, etc. There is no vetting of the expert advice as to the efficacy for solving a specific problem. That’s just what some expert said they used.
- Follow your friends. This methodology follows the advice of those around you because they liked it, or they have had success solving some problem.
- Follow experience. This methodology uses what we have always used within our current team experience. For example, if someone on the team has used some No-SQL solution in the past, and your application needs No-SQL, then simply use what someone on the team has already experienced.
- Research different solutions. This methodology seeks advice from all of the above. Then an additional phase is executed where the different technologies are evaluated on the best fit for the problem at hand. Often more than one implementation will be successful. That is actually the core issue with this methodology. You can waste a lot of time trying to get to a specific solution, and constantly question yourself as to having made the correct choice.
- Reverse Selection. This methodology is the reverse of all others. It finds a technology that solves everything. Then all new projects are implemented using the new silver bullet that solves all software problems for all time.
- Quote and a Goat. This methodology was introduced to me by a friend. His company simply asked for an estimate on when a product would be completed. It then identified who would be at fault when the project failed. Everything else was left up for grabs.
- Throw everything out the window. This methodology simply writes everything from scratch. It’s surprising how often this is still performed today.
What’s your methodology? Did you have a methodology you would like to add? Leave a comment, or drop an email to btaylor@sswug.org
Cheers,
Ben