Hi Pizza,
Thanks for the input.
What I'm seeing is opposite of what I would expect (i.e. a CD boot reading from a flash drive uses LESS memory than CD boot reading pupsave and .sfs from HDD). Given wear leveling concerns, I might expect the flash drive to load everything memory, and for the HDD to not load, leaving a large amount of free RAM with the HDD option (as wear leveling would not encourage loading everything to memory). Likewise with the USB1.1 versus HDD speed -- I would expect Puppy to read the sfs file from the HDD, and the flash to load into memory. I seem to be getting the opposite behavior.
BTW, my 2 install methods are CD boot and frugal install (I have grub on one of my machines, and plan to put it on the other). I didn't want a full HDD install (it makes my dual, Tri, or whatever, boot options very easy -- likewise playing with multiple Puppy versions). So I do have .sfs and pupsave files sitting on my HDDs (I still run Win98 on several machines)
My real problem was with my "toaster", which only has 256MB, and a tower with 384 MB. Perhaps the problem is that all of my machines have an available and enabled swap partition (making Puppy believe that it has more actual RAM than it does). BTW, the "toaster" only has USB 1.1 (and being a "toaster", it cannot be upgraded), but the 384MB machine has a 2.0 card inside -- however the HDD should be faster to read than the USB 1.1, so it still doesn't make sense... (and for completeness, the Duron mentioned earlier has 512MB and USB1.1)
The 111M memory usage from a CD/flash boot would probably be fine, but when it heads towards the 290M level, my machines spend a lot of time swapping, which really dogs them down (so to speak...
I think that I had no problem with Puppy 4.21 on an old laptop (P333, 128MB, using a CD boot with the files on the HDD).
Since the RAM values I gave were for a fresh boot...as soon as I open Firefox, or whatnot, I find that my free RAM drops to 5MB or so, and then pageswaps completely dominate my CPU availability... You end up with several seconds of pause after clicking a button...
If you could recall the name of your pfix code, I can probably search for it. I'll give the pfix option a try and see if that clears it up...
Thanks!