On 21 Jun 2007 09:11:31 -0700, in bit.listserv.ibm-main you wrote: >>To carry this argument to it's logical conclusion, each application >>would have it's own OS!!!! >> > > >It would be better to say that each application would have to include much >of the operating suystem function within it. >That would be too much to expect, especially with complex OS functions like >I/O management, paging and networking (basically common functions wanted by >all applications). >And how would these subsystems be installed and managed?
Everything uses components. Virtual machines such as the ones running Java might expand - but normal operations don't need to know much about peripherals. I suppose it is useful for high-end graphics to know what computing they can off-load. But whether the application knows whether the printer knows post-script can be variable. Virtual machines such as Macs running Windows or Unix are closer to the hardware level. But Intel Macs running applications designed for earlier chips show that this also is a variable. The dividing lines between OS, applications, peripherals, and remote computing are getting vaguer all the time. Just as we don't care what OS is in our router, we won't care what OS is running our data warehouse. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html