Re: Ada GC (was about Java VM)

Chris says


"I'm sorry for your bad experiences. I object to statements like
"abundantly clear" and "most experienced practitioners". I also don't
believe that the design phase of a programming language is the moment
when real experiences about ral time or performance are made.


In my experience good use of garbage collection (e.g not making stupid
allocations) results in code more efficient then using manual
allocation."


It is not a matter of bad experiences, just that no one was pushing
for GC during the design phase of Ada 95. We consulted a LOT of people
during this period, and far from their being a consensus to require
GC, there was just ZERO interest in this issue.


The issue in hard real-time is of course not efficiency but
predictability and removal of non-determinism. That's a little bit
different. I agree that for general programming, GC is highly
desirable (it is for me a fundamental dividing line between low and
high level languages).
--

Return to the comp.compilers page.
Search the comp.compilers archives again.