Because AGILE allows companies to commit SOX accounting fraud companies are adopting it in droves. Because it is yet another name for hacking on the fly without a plan kids, not professionals, kids, love it.
No. Getting paid to do something doesn’t make you a professional. Winning $20 shooting hoops in the park does not make you a member of the NBA and neither does getting paid to hack on the fly without a plan.
A bucket of user stories for the current sprint does not in any way shape or form qualify as a plan.
Mega Failure #1 – HealthCare.gov
Some of you are probably too young to remember when Healthcare.gov first went live. It was a catastrophe. The fraud masters fed stories to New Yorker, Washington Post, and MedCity News proclaiming HealthCare.gov would not have failed had it employed AGILE methodologies. To date I’ve not seen any print a retraction. Certainly not a retraction which gets echoed across the Internet. I must admit the old links to MedCity’s article no longer work so perhaps they fell on the sword that much.
Government Computing News called these bastions of journalism out. HealthCare.gov failed because of AGILE. What the proponents of AGILE fail to admit is that systems exist which are too big to AGILE. Some of those systems fit in your hand or into a small box hanging on the wall.
Waterfall was created for a reason. You have to define the scope and direction of a project. You can’t just hack on the fly until the money runs out which is the current Silicon Valley Startup mentality. Waterfall is how you can get a team of people to walk from Chicago, IL to Kansas City, MO without them drowning in Lake Michigan. Agile lets them drown and you iterate with a new team.
Mega Failure #2 – 737Max
I haven’t physically confirmed this and most importantly, nobody is denying it. You can do a Web search for “Boeing c++ agile” and, depending on the time of year, find Boeing with a lot of “agile” developer jobs. It also appears they’ve moved quite a bit of IT to India.
I have speculated on this in a March blog post. More and more are starting to support it. Some very familiar with crash investigations are supporting the idea that a stall control system which did not allow the yoke to override it per industry standard is a red flag of AGILE. When you don’t have a system architect and you don’t have The Four Holy Documents written up front, you end up with a catastrophe like this.
Hacking on the fly to a bunch of “user stories” is not software engineering. It is so far from software engineering that it cannot even mail a letter to software engineering.
Mega Failure #3 – The 2020 Iowa Caucus
We don’t even have to speculate here. Kids hacking out phone apps love AGILE. It lets them hack on the fly, write their own tests to prove their code works, and makes them feel professional. Too bad they aren’t professional and the tests rarely prove anything. Most are just there to check a box.
Too big to AGILE can be a system small enough to fit in your hand. iDiot phone developers never realize that. (The “i” in iPhone really stands for iDiot. Who else would spend a thousand dollars or more for a few hundred dollars in parts.)
An independent test team working from The Four Holy Documents when developing an overall test plan would have found this “coding error” that still as of 4:29pm on February 5th can’t give us the final numbers.
The Four Holy Documents
- Business Requirements Document (BRD)
- System Requirements Document (SRD)
- System Architecture Document (SAD; a.k.a. System Architecture Specification or SAS)
- System Specification Document (SSD; a.k.a. Functional Specification; or System Functional Specification – SFS; or System Design Specification – SDS)
It is time for the Federal government to ban AGILE in all industries. While the Iowa Caucus may be found to be funny, the 737Max and HealthCare.gov certainly weren’t.