Vitaliy wrote:

> I'm not sure why it's apparently so confusing, but I know I said we
> use *both* kinds. There are the "create once, destroy never" objects
> and there are objects that get created and destroyed numerous times
> throughout the life of the program.

Well, you didn't say it in the discussion about predictability. Your
argument was that the "create once, destroy never" technique is just as
predictable as static memory. My argument is that the lack of
predictability with the common use of heap memory ("create and destroy
all the time") is a problem with embedded systems that may not fail
under conditions per spec and that have little memory headroom for that. 

Your systems (at least the ones of the "create and destroy at will"
type) seem not to fall into that category -- IIRC, you said earlier in
this discussion that they may fail, and then just display an error
message to the user. This is something that's not adequate for many
systems.

The fact is that using the heap in the "normal" way makes memory usage
quite a bit less predictable than using static memory. For a certain
class of embedded systems this makes the heap a bad choice.

> I want to reiterate the fact that in some situations using SMA is not
> a viable option, so in those situations talking about its merits is
> completely pointless.

In the same condescending tone you're using sometimes I could respond
here that you probably just don't see how it could be done with static
memory, but I thought carefully about this and decided not to say it.

Gerhard
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist