The ICEman cometh. Yes, Peter I have inserted an ICE backwards and blown it, fertilizer happens, but that is not a reason not to garden. I have, on several occasions, been forced to work without an ICE since the client would not buy one. On all those occasions, It has cost the client more for time (mine, horray) than the cost of an ICE (e. g. CEIBO ~$1000). I finally got fed up and bought my own ICE. I am currently working with "the design from hell" which is someone elses "proof" that every rule regarding software and hardware can be broken and the result can still work (which in this case is true - most of the time). This design has 2 uCs working together on a shared bus using the most convoluted scheme and hardware imaginable. It takes me days to find even the smallest glitch. The 2 uCs must both be started exactly the same time and thus ICE is impossible. I have redesigned the system and will soon be debugging the project with 2 uCs, 2 ICEs, 2 PCs and having a very productive time even using most of the existing crummy software in order to meet the deadline. I do agree that some means of debugging is possible with code changes and LEDs, the problem with that is that when you find out that the point you chose was not "it", you have to go through a full edit, compile, download to establish your next checkpoint, with an ICE, you just set another breakpoint. A point raised that you can not always believe your ICE, true, but is there any tool you can always believe?.

Have fun, stay ICE cool

Erik