Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!rutgers!apple!oliveb!amiga!cbmvax!jesup
From: jesup@cbmvax.UUCP (Randell Jesup)
Newsgroups: comp.sys.amiga
Subject: Re: 1.3 bugs
Message-ID: <7877@cbmvax.UUCP>
Date: 11 Sep 89 23:46:27 GMT
References: <26736@shemp.CS.UCLA.EDU>
Reply-To: jesup@cbmvax.UUCP (Randell Jesup)
Organization: Commodore Technology, West Chester, PA
Lines: 25

In article <26736@shemp.CS.UCLA.EDU> michael@maui.cs.ucla.edu (michael gersten) writes:
>Here are some bugs in 1.3 that I'd like to see fixed for 1.4.
>#2. Although you can "assign remove" to unmount a device (why not mount
>remove?), you do not recover the space used by that device. How about
>a new call vector for devices corresponding to expunge in libraries?

	It's trickier than that.  The killer is that someone who DeviceProc()s
a handler can hold that msgport address forever before using it.  Solution:
come up with a way to say "I'm done with the msgport".  Works, but old
programs that call DeviceProc directly will cause problems.

>#4. (You'll love this). Mountlist needs a new field -- DriveReserved.
>I have a nice, high performance drive that reserves track 0 for its own
>activities. This gives me 14 tracks on cylinder 0, and 15 on all the other
>tracks. But if I say 15 tracks per cylinder in the mountlist, it does
>inefficient seeks all the time. (And if I say 1 track per cylinder, it
>is less optimal in file layout and slightly slower).

	That's easy.  Increase the Reserved= number to account for 1 full
track.

-- 
Randell Jesup, Keeper of AmigaDos, Commodore Engineering.
{uunet|rutgers}!cbmvax!jesup, jesup@cbmvax.cbm.commodore.com  BIX: rjesup  
Common phrase heard at Amiga Devcon '89: "It's in there!"