Pretty funny - the scenario you describe is pretty much exactly what happened when I inherited the management of a site that gets about 257,000 uniques per day

Pretty funny - the scenario you describe is pretty much exactly what happened when I inherited the management of a site that gets about 257,000 uniques per day - millions per year. It was built using Rails 2.0 without any attempt to upgrade it since it was first deployed. Since I really don't have time - my time is taken up managing 3 to 4 to 5 other projects at the same time as managing this web site - I've been unable to spend the pretty large amount of time it would take to bump the application up two major versions so that incremental updates become easier. Nor does the company want to pay me to do that. I think a lot of applications reach an early end of life because of the nature of subcontracting. You build the application, your contract reaches an end, and then the company wants to stop paying for upgrades, seeing them as optional. They will only pay you to keep administering the existing site indefinitely. But where I might differ? Maybe that's the way it should be, short of being reckless (you should always be applying security updates). Technology and methodology changes so much in three or four years that retrofitting new features on the old, crufty application seems like going backwards. Ultimately the company ends up wanting to rebuild the whole thing. In the case of the application I'm talking about, almost all of its features would be done better by an existing CMS system. I would feel sick to my stomach adding new CMS features that I know would work more beautifully inside of off-the-shelf CMS software that has been extended to complete the custom task.