Should you correct production mistakes?

On Wednesday (08/25/2010 at 10:59PM +0100), Tony Duell wrote: _I_ would fix it. It's clearly not what HP intended, sadly it slipped through some quality inspection and certainly isn't supposed to be that way. I'd take the position of fixing it now, while the unit is apart for other reasons, so that you don't have to take it apart again later when it does fail and then risk damaging other aged things then. I had a similar decision which you folks offered input on-- which was the apparently failed transformer in my SWTPC 6800 system (c. 1976). I failed to close the loop on that story but in the end, I discovered that the real problem was the lousy MOLEX connectors that were used between the power supply PCB and the end of the power harness going to the motherboard. The pins in this connector were not mating well, were corroded and had been heating up for some time (a long time ago!). The white MOLEX shell was turning brown and was actually warm to the touch when I would briefly power the machine on. I decided to replace this connector with something similar but better and all the issues went away. Chris -- Chris Elmquist