Hi There has always been several problems with the central repair shop method. The biggest is lost expertise. When done like this, only a few people understand a particular module. With people moving on to other jobs, deaths and other things, sometimes the needed knowledge gets lost. It iskind of like maintaining a minimum population of a species. There is another problem that I have had personal experience with. When the module it sent back, the test environment isn't the same as the real one. They find no problem with the module ( sounds like the car thing doesn't it but this is a different case ). The module is sent back with the problem still in it. This was from a manufacture and the same error would occationally happen on the line. At one point 15% of the units received were experiencing the same problem. A quick check on serial number showed that most all were returned for that problem. It eventually got straightened out but I can tell you that it was annoying to turn on the computer and find that the chassis was wired to the hot lead of the power. You'd ask your self how could such a problem exist without someone at the factory noticing it? Although this was a specific case, the problem is generic to the of assuming that any test, other than the one in the real application, is truely a valid test. The more we depend on the specmenship to define the test criteria, the greater the risk that things like black boxes in airplanes will fail. Things are just getting too complicated and there is thought that the DEPO concept can keep up with it. My experience has shown that it just makes the problem harder to deal with and the reaction time, in case of a problem, to be slowed by red tape. Dwight On Thu, 8 Mar 2001, Steve Robertson wrote: