Geek: healthcare.gov source code...

Geek: healthcare.gov source code...  Weeks ago I started seeing news sites talking about healthcare.gov having 500 million lines of code.  That's a crazy number, way higher than any reasonable estimate, and I didn't bother to repeat it as I thought it was likely bogus.  Now we have a report from a programmer on the project, saying that there are 3.74 million lines of code underlying healthcare.gov, and offering a breakdown by code type (at right).  That's a far more believable number!

While one could debate whether CSS and XML are “programming languages”, it's certainly true that all the things listed are the typical work product of programmers on a project like healthcare.gov.  The ratios look about right, too, except that I'm a little surprised that the JavaScript component is so (relatively) small.  The dominance of Java on the server side is entirely typical of a large corporate project these days.  I'm somewhat relieved by the absence of Basic, .NET and C++.  I'm also a little surprised that a government contractor didn't manage to sneak in some proprietary language (like CMS) or some off-the-wall or obsolete language (like Forth or Prolog).

None of the above should be taken to imply that the healthcare.gov software was well-specified, well-designed, or well-implemented.  In fact, I think we can assume that none of those things are true, given how badly the system bombed on rollout, and how much of it remains to be implemented.  All this information does is to say that the implementation isn't completely, overtly, insanely crazy...