Obamacare Fiasco Intentional?

5 The Market Ticker by genesis  /  40min  //  keep unread  //  trash  //  preview The Market Ticker - Obamacare Fiasco Intentional?   +TAG Forbes has an interesting spin on the Obamacare fiasco, and it may be the right answer: A growing consensus of IT experts, outside and inside the government, have figured out a principal reason why the website for Obamacare’s federally-sponsored insurance exchange is crashing. Healthcare.gov forces you to create an account and enter detailed personal information before you can start shopping. This, in turn, creates a massive traffic bottleneck, as the government verifies your information and decides whether or not you’re eligible for subsidies. HHS bureaucrats knew this would make the website run more slowly.But they were more afraid that letting people see the underlying cost of Obamacare’s insurance plans would scare people away. In other words, they knew that the principal claim used to sell Obamacare in the first place, that costs would come down for the majority of people, was a flat, bald lie. So as with virtually everything Government does, instead of telling the truth they decided to intentionally lie, that is, defraud you, even if in a "soft" way.  Why?  Because there is never any penalty assessed against any government official or branch of the government when it lies and screws you.  Ever. The rest of the problem lies, as I have noted in a few other posts and as have others, in the choice to use "asynchronous" data collection from various places.  This technique is commonly (ab)used to try to make pages look "snazzier" as data "fills in" since the pages comes up before the data is present.  The problem with it is that this technique is inherently dangerous when there are capacity constraints, as computers don't slow down the way an amusement park ride queue does. Computer systems respond to increasing load in an interesting way.  So long as their ultimate capacity is not exceeded the per-client service time rises but the aggregate transaction rate also rises.  But then the "knee" is reached and instead of the per-client service time going upit effectively collapses. One of the tricks in good software design is to detect these points early, before they happen, and either graciously refuse service up front or degrade the service in some way to avoid the knee from being reached.  The forum here, for example, responds by engaging in various load-shedding operations, including removing the user avatars from displays -- the most-visible of the change in response.  This allows the total transaction count to remain high and the number of clients served to continue to grow, even though the system is under extreme load, because it sheds off some non-essential functionality. But as soon as you decide to force people through a process that is multi-threaded and has a lot of various knee points, none of which you directly control, that goes right out the window. There is one other point to consider -- Obamacare's web system may be designed to force registration so if you don't buy a policy the government will have a ready-made list of people to come after in the future to collect their fines. After all, you did provide all your personal information first, before you saw the prices -- right?