Hi Erik,

Well, that is a reason. And frankly, I find it more refreshingly frank than the typical "white paper" response wherein someone goes through a litany of bullet points and logical contortions to try and justify such a position, without ever admitting their real reason.

I used to have a similar position on IBM. I've said before that my first computer was an IBM PS/2 Model 50. IBM marketed their PS/2 line to college students (though I traded a very expensive rifle for mine from a friend). But when my 20 MByte hard drive filled up, I found that the only way I could get one with more capacity was to give IBM $1600 for an MCA (Micro Channel Architecture) 40 MByte drive. Any normal 50 MByte hard drive would have cost less than, or not much more than, $100 at the time. When I bought a pen plotter and needed to add a serial port, instead of getting a $20 card, I had to buy a $125 card. Eventually I got tired of haveing to compress one directory so I could expand the one I needed, but instead of giving IBM $1600 for another hard drive, I bought a new 386/387 machine for $1000 (with an 80 MByte hard disk. Wow! I'd never be able to fill that one up :O ).

In a recent post I mentioned writing an ion source extraction model in BASIC because I couldn't get a Fortran compiler. In point of fact, the reason I couldn't get a Fortran compiler was because the only one in the local store at the time was an IBM compiler, and I wouldn't buy it. I did order a non-IBM Fortran compiler, so I didn't do any more models in BASIC, and eventually I learned enough about C/C++ to do what I need. But to this day I have never spent another dime on anything IBM.

As for Atmel, I use Atmel routinely and will continue to do so if I can buy them from Mouser. But then, my experience with their field engineers in Dallas was pretty much the diametric opposite of the experience you describe.

Later,

Joe