Date: Mon, 22 Apr 1996 13:29:37 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To: Jih-Shin Ho <u7711501 AT bicmos DOT ee DOT nctu DOT edu DOT tw>
Cc: djgpp AT delorie DOT com
Subject: Re: Detecting drives
In-Reply-To: <4lf1hj$mhm@ccnews.nctu.edu.tw>
Message-Id: <Pine.SUN.3.91.960422132254.16999B-100000@is>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


On 22 Apr 1996, Jih-Shin Ho wrote:

> : Not so fast.  "Network drives" might mean Novell 3.x, and you'd be 
> : surprised to see what you can get there, since Novell 3.x hooks Int 21h 
> : before DOS and thus any info about Novell drives bypasses DOS entirely.
> 
> If DOS doesn't understand this drive, how can it use the drive ?
> For example, 'dir x:'

Like I said: Novell hooks Int 21h and therefore gets to see any DOS
requests before DOS.  If it sees that the request references a disk that
is handled by Novell, it services the request, and DOS might not even know
that such a request was issued. 

> But the title is 'detecting drive', not 'detecting valid disk in drive'.
> 4409 only tells you which drive is valid, it can't tell if there is a
> disk in drive.

You don't detect drives as an end in itself.  You detect a drive because
you want to do something with that drive after you detected it.  What I
wanted to warn about is that when you issue DOS calls that refrence CD-ROM
drives but don't actually access the disk, you could think the drive is
valid, but when you then try to access that drive, you get GPF.  For this
reason I think a drive without a disk should be reported as an invalid
drive, or at least some indication of a possible problem should be 
available to the caller of such a service.