With COM port speeds, neither heap fragmentation nor memcpy will have any
significant effects.
BTW - memcpy is a worse penalty then fragmentation, since the kernel pool
allocation algorithms are rather smart. If your allocations are all of the same
size, then use the lookasides for them.

Max

"nospam" <nospam@nospam.nospam> wrote in message
news:Xns93CBB8AA8B919nospam@194.90.1.17...