Mail Archives: opendos/1997/05/30/06:16:35
On Thu, 29 May 1997, Steven Ehrbar wrote:
> > > This might sound silly, but why not set up the system so that it first
> > > checks for the filename in the case you typed it, and if that fails
> goes to
> > > the first name in that directory FAT that matches case-insensitively?
> <snip>
> > The old version in \ASDF is deleted. I then forget all about the
> > whole thing and use the new asdf for a couple weeks. One day, I
> > decide to clean up my hard drive because I'm running out of space
> > so I go looking for things to delete. Then it occurs to me that
> > I had an old copy of asdf laying around in the \ASDF directory.
> > Did I delete it allready? Hmmm, I'm not sure so I'll just delete
> > it again and be sure:
> >
> > DELTREE \ASDF
> >
> > The heuristics engine can't find \ASDF so it looks for other
> > filenames with the same letters in different case, then it
> > deltrees \asdf and I no longer have my asdf program.
> >
> > No thanks.
>
> So you throw a confirm into the delete and deltree commands that warns you
> when you're violating case and gives you the option to back out, similar to
> the "del *.*" confirms. So then the user sophisticated enough to properly
> use case-sensitive file names runs a "dir" to see what's going on, and the
> others suffer the same fate as unsophisticated "del *.*" users.
If you throw a confirm into it, you could still accidentally hit
enter and toast something, only then realizing and being unable
to stop it. Also, what happens in a batch file? For example:
@ECHO OFF
DELTREE \ASDF
In this case, the batch file may need to be ran unattended, such
as on a BBS, or something. A batchfile that hangs expecting user
input could bring down the BBS, or cause countless other
problems.
> > > And, of course, make the whole thing (whether you want total case
> > > insensitivity, simple case retention prohibiting case-differntiated
> files,
> > > or full case sensitivity which allows case-differentiated files) an
> OPTION
> > > in the config.sys file -- maybe even a per-filesystem or even a
> > > per-partition option.
> >
> > As it stands with operating systems we currently have enough
> > problems to worry about rather than implementing a bunch of new
> > convoluted systems.
>
> > 1) Case insensitive: DOS/VMS
> > 2) Case insensitive stored case: VFAT/NTFS/HPFS
> > 3) Case sensitive: ext2/minix/other unix FS's
>
> At the time I wrote that (literally /weeks/ ago!) nobody (at least on the
> non-developer list) was talking about limiting case sensitivity to non-FAT
> file systems. Accordingly, my config.sys proposal was primarily dealing
> with a modified FAT system, and was modeled after the options in the
> config.sys for OS/2 for ext2fsOS/2.
Well, we *have* talked about leaving FAT the way it is though.
Any improvements/changes would be a SEPARATE filesystem, perhaps
*based* on FAT (I can't think of a single good reason why
though).
What is ext2fsOS/2?
> And people who edit a config.sys file without knowing what they're doing
> deserve trouble, while people who do know there might be filename conflicts
> when they switch from sensitive to an insensitive option and will be
> prepared to deal with it.
My proposed IQ test would circumvent someone from choosing
options that could cause problems. :o)
Mike A. Harris | http://blackwidow.saultc.on.ca/~mharris
Computer Consultant | Coming soon: dynamic-IP-freedom...
My dynamic address: http://blackwidow.saultc.on.ca/~mharris/ip-address.html
Email: mharris at blackwidow.saultc.on.ca <-- Spam proof address
URL: Sun Microsystems http://www.sun.com
- Raw text -