Dunno...maybe I'm just lazy (or stupid) but starting with XP I just set a fixed pagefile size to the maximum amount the system automatically determines when it's allowed to manage memory. And I have never
experienced an OOM situation doing it that way.
As Stoic pointed out earlier, Windows is
actually very good at managing it's memory space. So unless you have some very unusual requirements, most tech voodoo and deviation from the basics seldom nets a benefit worth pursuing. And with an operating system as locked down (and with so many undocumented subsystems) as Windows, it's not like you can always know what your mucking around with system and low level settings will do.
To me, life is far too short to bother with most of that unless I need to fix something that's broken. And most times, throwing in some additional RAM accomplishes the same thing faster and better anyway.
Ramdisks are another story. I use them (on servers) for rapidly updating temporary file caches and logs. But that's not something most people (except engineers or graphics pros) would ever need to worry about on a workstation. My feeling is that if you really do
have a legitimate use for a ramdisk, you already know when, where, and why you need to set one up.
The people who do America's Test Kitchen
magazinehave a philosophy I apply to much of the system tweaking I do. They're always asking." What does the extra work get you?" When Cooks
publishes a recipe, they try out every variation (ex: 'milk heated' vs 'at room temperature' -or- should you use plain yogurt or
sour cream) they can get their hands on. When they're done (they once tried 35 different recipe variations for sugar cookies!) they can tell you exactly
what matters and what doesn't. What's a legitimate concern, and what's just old-wives tales.
Some of the most interesting and beneficial comments come when they find something fussy and persnickety that does
make a difference. Because at that point, it becomes necessary to decide if it's worth the extra time and money. One recipe called for some extra steps and a seasonally hard to find ingredient. Their conclusion? It was the best recipe - but not really worth it for the added expense and inconvenience it entailed. The runner-up recipe gave the taste testers 99% of what the ideal recipe did - but with considerably less work and fussing. Conclusion: go with the runner-up for most occasions.
I approach system tweaks and optimization the same way.
In my field of business there's a saying: The first 90% of a project consumes 90% of the budget. The last 10% of the project consumes and additional 90% of the original budget.
I found that to be pretty much the case.
Just my tuppence.