|
From: Alain <al...@po...> - 2005-06-29 18:08:55
|
This was private, bu I forwarding to the list. I believe it can be of general interest. The list is "devel" after all... Hi... > In fact DOS=HIGH also makes it slow too, but only around > 10 to 20%. Expected. It puts BUFFERS into HMA but avoids to access them directly there, so you get one extra "copy to/from low RAM" step. Still DOS=HIGH is really nice. > It is the same with UMBPCI or EMM386. > Loading UMDA low, or the /L switch makes no difference. Very strange. How about not loading ANY drivers into UMB and just enabling UMBs in general (DOS=UMB, umbpci/emm386, but no LH or devicehigh or similar...)? > > Even more: You could make calccach available for others :-)). > I believe it is of little interest no anyone... You are, as said, not the first to ask about some automatic way of selecting cache sizes, so others should be interested as well. Would be nice if you could exchange some ideas with Erwin to come up with a wishlist of features that calccach can have in the future. > >>>Alternatively, calccach could just start the cache itself. > >>I opted for the safer way, if the program dos not run (i really don't > >>trust my users) I can use a default value. > > > > That would only happen if the users modify their autoexec, and in > > that case, nothing would help anyway. And: If it does not run, it > > does not have an errorlevel either. > > "Users can do the unexpected" As said, if you have calccach lbacache %errorlevel% you have exactly the same "protected from stupid users" level as with calccach lbacache ... Eric |