On Tue, 6 Jul 2004, Arkady V.Belousov wrote:
Hello,
config.c - config now parsed in 5 passes:te> that's fine IF config.c fit's into buffers or you boot from disk. te> else (if boot from floppy) that might become *SLOW*
There was 3 passes, now 5. Not a big change, especially because on some passes there will be other disk accesses (drivers, installs, shell). When I test config.sys working (on floppy), for me this is enoughly fast. Of course, my config.sys-es are much lesser than 10k (20 buffers are preallocated before processing config.sys), but I don't think that you will notice "*SLOW*" even on bigger config.sys.
But what you propose else? 5 passes are _required_.
What I still wonder about this topic -- I remember that I had a CONFIG.SYS size problem with Win95 (when its size grew past some point, DOS 7 did't start) -- aren't device drivers allowed to mangle with drive mappings? I mean there is (was?) a DSWAP.EXE or SWAP.EXE or whatever from stacker that allowed to "compress" the boot drive by swapping C: with the drive letter the compress-on-the-fly drive got. This was before MS came up with the DRVSpace.BIN approach.
I actually believed that CONFIG.SYS is to be kept in memory because of this stuff. Is this handled silently by the buffers?
Bye,
--
Steffen Kaiser
-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
Freedos-kernel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/freedos-kernel
