Once someone attempts to do what I've described, he/she will quickly see that it may get around the limitation on size or location of memory in which the code created with the compiler will operate, but it still won't fully enable the functions of the other tools in the tool suite. If one wishes to do "serious" work, it's not really applicable, but ... given that one is determined to try out the tool set with a device with which it won't otherwise work at all ... since there's room for doubt as to whether the trial version was ever intended ot produce linkable/runnable code at all, though it may do so, it merely allows the user to do what the evaluation version ostensibly does allow, namely to give the tool set a more thorough trial, albeit with a very small memory.

I've never been willing to do this, but if one were willing to put forth the effort, it might provide a way to try out the compiler under conditions not forseen by the authors.

I understand why persons whose primary product is software/firmware might object to such practice. However, anyone, particularly one who's oriented more to the hardware than the firmware, who's purchased a software product only to find that it didn't "measure up" will probably see the method in my madness. If the "demonstration" version won't demonstrate what you need to know about a software tool suite, you sometimes have to beat on it in order to make it do what it is supposed to do. Sometimes that actually works! Don't forget that the guy who's likely to want to use a compiler in the first place is probably more interested in using the "real" tool rather than hornswoggling the demo version. Moreover, he probably wants to give several competing tool suites a bit of a workout before deciding which one he wants to live with.

RE