wylie@cui.UUCP (WYLIE Andrew) (09/23/86)
Hello,
I am having some problems with the second ram disk (RAM.SYS + SETRAM)
which was recently posted to the net. This has raised some questions about
disk drives in general which I hope someone out there can help me with.
The problem I have with SETRAM is that SETRAM d N works, but SETRAM e N,
or any drive greater than d:, gives the error
error number 15 (decimal)
so that I can only create one ram disk. (I shall ignore the temptation to
elaborate on the wonderfully informative nature of the error message).
Now, if I do a simple DIR d:, I get the message
Not ready error reading drive d
whereas DIR e: gives
Invalid drive specification
I assume that this is related to the fact that the diagnostic diskette tells
me that drive d is not installed, but says nothing about drives e,f,g... etc.
All of this is completely unchanged if I set LASTDRIVE=Z in CONFIG.SYS
The questions I have are :
1/ how does the diagnostic prog think that drive d exists (but is not installed)
but that higher drive numbers do not ?
2/ does LASTDRIVE actually do anything ?
3/ can I make the PC think that more drives exist but are not installed ?
4/ will this cure the ram drive problem ?
Andrew Wylie, Univ. of Geneva, Switzerland
---------------
mcvax!cernvax!cui!wylie
WYLIE@CGEUGE51.BITNETbet@ecsvax.UUCP (Bennett E. Todd III) (09/29/86)
The diagnostic messages are fine for what the program was written for --
as a utility for the author. Redistributing it was my idea, not his; I
thought other people might find it useful, as it is so amazingly simple,
and works so well. Oh well, no more.
If you want more than one RAMDISK then put more than one line (one for
each disk) in CONFIG.SYS. They will get consecutive letters. Sometime
you might need to bump the LASTDRIVE specification; I am not certain
about that one. If you want to play with multiple RAMdisks, beware
especially about not making holes in DOSs memory map (allocate and free
your chunks of memory in strictly stack order) as warned in SETRAM.DOC;
we haven't actually seen any problems from such segmenting of RAM, but
that does exercise a part of DOS which isn't normally exercised, and so
can be assumed to have fatal bugs.
-Bennett
--
Bennett Todd -- Duke Computation Center, Durham, NC 27706-7756; (919) 684-3695
UUCP: ...{decvax,seismo,philabs,ihnp4,akgua}!mcnc!ecsvax!duccpc!bet