On 23 Jan 2005 at 1:59, A-NO-NE Music wrote: > David W. Fenton / 05.1.22 / 07:37 PM wrote: > > >Mac users are probably confused about virtual memory because of the > >way it was bolted onto Mac OS as an extension that often conflicted > >with applications. > > > >You're now living in a completely different world, one where the apps > > don't know anything about managing system memory at all. They only > >know about their own memory requirements. > > > >Keep in mind that under OS X you don't have to set the amount of > >memory to be allocated to an application (as was the case with Mac > >OS). This is a direct indication of the fact that memory management > >is now wholly owned by the operating system itself. > > > I don't think you understood me. You seems to think I was talking > about Classic Mac OSes, while you are clearly based on NT > Kernel. . . .
Eh? What are you talking about? Windows NT? What does that have to do with Mac OS? I'm talking about OS's in general, not about the specific architecture of any particular virtual memory manager. > . . . Did > you know OSX blocks swapfile allocation in 67.1MB each for first two, > then size multiplies after, i.e., the 3rd is about 135MB, and the 4th > is about 268MB, and so on.? It's not like NT Kernel does at all. Only in implementation details is it different (and I'd wager that the increment size is something that can be changed). In the basics of interaction between VM and applications, it's pretty much the same: applications can't move while the VM is paging. > And, as I said before, there are some apps that calls vm at launch, > and the vm size never changes after that regardless of the presence of > available physical memory and the size of the project file. Applications don't call the VM. The OS does, based on the memory requirements of the applications the OS is running. > And to your other post, as far as I know, OSX did not take vm > architecture from BSD kernel, which was the base design of Darwin > Kernel. But virtual memory subsystems have certain things in common or they wouldn't be virtual memory subsystems. One thing that is common to all real VMs is that applications don't even know they exist. -- David W. Fenton http://www.bway.net/~dfenton David Fenton Associates http://www.bway.net/~dfassoc _______________________________________________ Finale mailing list [email protected] http://lists.shsu.edu/mailman/listinfo/finale
