Ah. Witness the phrase 'shouldn't be that hard'. It is not fair to the OS3 programmer as he a) has to port the software first; b) hope there isn't anything too UNIXy deeply buried within the code---I noticed that newlib supports sbrk() which cannot be properly implemented on OS3 for example; c) hope that the implementation features are the same on OS4; d) do the work himself even though other contestants may be doing the same thing; and e) has to get newlib working in a similar way as it does on OS4. Which is a mite tricky, as newlib on OS4 seems to be a dynamic library, while my first impulse would be to code it as a linker library. Perhaps I am mistaken, I do not own OS4 nor its SDK. If I am right, the OS3 programmer has a complex job on his hands because he must also write his own startup code for the compiler. My, what fun.
All in all, the OS3 programmer has to jump through hoops in order to get the code working, and all that effort is essentially wasted since the goal is the development of OS4 programs, not OS3->OS4 glue code. Which makes me question the reason why on earth you would want to encourage OS3 developers in the first place: OS4 is not OS3. It's like developing OS3 software on the OS1 platform: doable, but masochistic at best. The fact that the winner can now choose between money or an A1 is a good change; the next change should be that you in fact do not recommend people developing on OS3 at all, or that you create a second category specifically targetting them. The prize would of course be smaller. It clears up the contest rules to a considerable degree, and levels the playing field.
I am very sorry to sound so critical about what at the core is a great idea; I just get the nagging feeling that the implications of the rules haven't been properly thought through, something which other people on this site noticed as well.
