You imply that to write an application takes significantly shorter time in C than in asm. I say, this depends. Also, I say that there are other options. Interpreted BASIC, for example. Or, for that matter, said Pascal - and if you are concerned about bugs and upgrades, bribe the author for the sources ;-) Or, rolling out your own RAD tool - why not, if it fits the situation.



Sure. This is a very common practice today. Again, I say, it depends, whether it's worth messing with C at all, or the other way round, whether it's worth messing with asm at all. There is a whole host of applications and there is no one tool fits all, although the toolmakers would like you to believe that. A good engineer should have the toolbox filled by a wide variety of tools - yes, including the sledgehammer.



Asm code can be reused, too.

During the years I have ported a lot of my programs between different architectures, asm to asm. You can see a snippet of that on that on my website - the block cipher implementations were written on '51 and rather straighforwardly ported to AVR. It was never a "downport" to a less capable processor/core/system, but I doubt that would be the case of your ports either. Most of the stuff was easy and went quickly and the hard stuff would be hard in C either, as it was the machine specific part anyway. What you spare down in C is the raw typing, but that's almost never the real issue.

So, I say, the portability advantage of C is a myth (not because it is not portable, but because it is none more portable than asm).

JW