Today I tried to load the latest changes in an assembler application program into my Mega168 AVR. Like I have done a hundred times before in the past few months using Studio 4.
It didn't work. I got an error saying that the Studio program could not connect with the programmer. This is a Pololu standard ISP programmer that has a USB to serial port function and a serial port-to-AVR programmer. It has *usually* before in the past. Like all USB junk, it fails about one in ten times of ordinary usage. Usually I pull the cable out of the USB port on my PC and plug it back in. Then it *usually* works.
Imagine that your car worked like this: the automaker announces that they have a new 'engine' that makes the car go ten times faster. You don't really need the speed, but what the hell... However, there is a catch. In one out every five or six times that you try to start the car, it just won't start. For no discernible reason at all. But if you just get out, take off the left rear tire, and then put it back on, it will probably work now. Again, for no discernible reason at all.
Nobody would accept this situation from an automobile, but it's standard operating procedure for an AVR programmer. WHY?? WHY can't Atmel just fix this so that it 'just effing works'? Why is it so difficult for Atmel to make JFW quality software. JFW of course means ..just effing work... Your car is JFW; your computer mouse is JFW, your AVR programmer is a lot simpler technically than a car or a mouse, but it has never been JFW status software. JFW has many names, some people call it 'six-sigma', some dare call it 'quality', some don't believe that the computer/electronics industry will ever be able to deliver JFW for any product.
So, anyway, since I (like MANY people on Earth) don't have internet access at home, I went to the local library, plugged in my flash disk doo-dad, and downloaded all 750 MEGAbytes of the new (but in now way improved) Atmel Studio 6. Then I started the loader on my home PC, and waited a whole G-D hour for it to load .NET SQL and a million other pieces of junk software that I had no idea that were absolutely necessary to load a hundred bytes of code into a Mega168. When Studio 6 finally finished loading, I tried to load MY assembler code into it an *hopefully somehow* get the source code into hex code and then into the Mega168.
No such luck. All the effing documentation either has to be downloaded from the Atmel website, or, it is a .mshc file. Now I use this PC for writing and loading source code into an AVR chip. Why would a PC used for this purpose need, want, or have internet access? Internet is very expensive here. Many of us just go to the local library to access it for free. So if idiots would make a 750 MEGA-effing-byte program that just compiles C source (and not as good as the MOSDOS Turbo-C at 50K bytes) and dumps it into a $2 microchip, then WHY would they not include a 20K file (in plain text ASCII if you please) that explains how to use it? Like they did ten years ago with Studio version 1?
Every year, every revision this whole situation just gets worse. I'm at the point where I wish that I could just get a cheap computer with Windows 98 that actually runs all the programs without BS that I used to run without BS when my home PC was running Windows 98. But I need a flash drive interface, so I guess I should be asking for Windows 2000.
The idea that every PC should run every program written for every PC Windows version at least 90% of the time is just crazy. I want one dedicated PC (or Raspberry or whatever) that I can bang out source code in AVR assembler, assemble it, and stuff it into an AVR chip. And I want it to work 100 effing percent of the time! I don't care if it uses an 'obsolete' DB9 or DB25 or 1940's phone plug connector. I just want it to work all the time.
I want a JFW-quality computer-to-AVR programming interface. I don't want another 200 versions of the AVR. I'm willing to use Mega168/328 forever.
I would also like peace-on-earth and goodwill-to-all-men. But like quality software from Atmel, it is just dreaming the impossible dream....