>Don't place too much importance on the CPU. The hardware I/O >involved in loading and swapping takes up the bulk of the time.
Of course. >wrt your Windows benchmarks: If you opened MS Word while Outlook >was still open, this can drastically reduce Word load time; as >components of MS Office, they share many DLLs. I didn't make that mistake, although I am sure I made others. I did try to be reasonably accurate by warm-booting both machines and waiting until they were done loading services and startup processes, so the applications could be launched from a relatively idle system state. >Faster hardware is a dangerous and poor substitute for efficient, >well-designed software. I have never witnessed a comp.lang.* >usenet thread to this end, though I have heard individuals state >this opinion. And therein lies my point. Programmers don't matter. Engineers don't matter. Computer scientists don't matter. Only unschooled users matter, and they don't care about graphs, arguments, explanations or qualifications. If things run faster for them, they don't stop to ask whether better hardware or more efficient software deserve the thanks. They do care about booting fast, getting their jobs done, using programs they already know (insofar as possible), and not having to watch an hourglass. Since you asked for a comparison of equivalent applications, how's this: Launching OpenOffice Writer on 1GHz Win2K machine: 12 seconds. Launching OpenOffice Writer on 2.4GHz Linux machine: 21 seconds. _______________________________________________ RLUG mailing list [EMAIL PROTECTED] http://www.rlug.org/mailman/listinfo/rlug
