LinkedIn
Software Architects and Enterprise Architects Group
How would you describe a well-architected system?
@Gary this is exactly my point of view (based on similar experience).
Taking one fabricated number as the key measurement (without questioning how effective it can be in most cases, in our economy today!) is dangerous. It hides the precision of the method creating it (not to forget about that not-really-Churchill-quote saying "I only believe in statistics that I doctored myself"), and the likely motivated reasoning behind. And it does not replace a conscious risk analysis for the future.
@Edward, I have checked your site, and we could argue for years without convincing each other, because we have totally different experience. I assume you have helped lots of companies with your strict rules as a consultant - I have seen the same rules breaking the neck of systems from inside, as an architect. I accept the importance of your concept with one exception: your statement of "no exception". I think the ROI analysis-based barriers can and should be broken _sometimes_, after conscious and responsible negotiation. The same applies to my side: the best architectural ideas can be put aside _sometimes_ because of the too risky ROI estimation - after conscious and responsible negotiation.
My mantra, the "conscious and responsible" is missing lots of times both on management (Holy ROI) and designer/developer side (Holy Architecture, Holy Tools, Holy Patterns, etc.) - and without an open communication, it becomes a power or marketing (polite way to say bluffing) battle, a poker game.
Yes, even conscious and responsible people can make mistakes - but at least they can identify the causes and make better decisions in the future, adding ROE (return on error :-) ) to the equation. To move forward, we _must_ fail several times before reaching a higher level goal, so in several cases we _must_ pay the price for nothing (but experience). "Experience is what you get when you don't get what you wanted"... I got this from Randy Pausch, a man worth knowing about.
If we all focused on measured and stable ROI, we would have nothing from our current achievements in science, technology or society. IT should be something like "controlled and monitored creativity"... No surprise that I work today on SW Quality Assurance, with concepts like: "the best code is _not_written_", and work literally years on a framework that actually minimizes the required amount of code :-)
@Gary, back to the example: this approach does not ruin America only, but threatens the entire world today. Space exploration is inherently dangerous, and we stall because we are blinded by Star trek-like movies (main heroes always survive) and afraid to lose precious human lives, though it is part of the real game.
On the other hand, we play with technologies that can really destroy countries, have global consequences, with the same shallow thinking, and _that_is_really_bad_. I think this lecture is worth listening to... http://fairewinds.org/podcast/fukushima-daiichi-nuclear-accident-ongoing-lessons
