Learnings in S/W Engineering from the HealthCare.gov website fiasco

It cost at least $400 million to put together the problem-plagued HealthCare.gov website. What went wrong?

The PM perspective (from Don’t Blame IT for Obamacare’s Tech Troubles):

Blaming programmers, coders, and project managers for disgraceful design flaws and technical turmoil is too easy and obvious. Crap rolls downhill. Look deeper. The underlying truth for virtually every large system’s implementation initiative is that success demands leadership and oversight that holds itself accountable for assuring best practice. Good governance, not superior technical chops or ready access to alpha geeks, is how you build complex systems that deliver reliable and resilient value for money. Good governance provides oversight, insight, and foresight. Emerging problems are flagged sooner; project leaders present their testing protocols and outcomes; and updated expectations are clearly communicated throughout the enterprise. Contingencies are constantly reviewed and appropriately revised. That’s what serious systems developers do.

Web Performance Engineering perspective

Top Performance Problems:

• Too many HTTP Requests (46)
• Overuse of JavaScript (12 files total, 166K compressed, 464K uncompressed)
• Use of large offsite web fonts
• Poor caching of resources
• Use of offsite third party widgets
• Slow TTFB times
• Uncompressed images (251K total, 75K could be saved losslessly)
• Use of SSL resources on non-secure page
• Unminified JS and CSS files

to be continued...