Conference hydra::amiga_v1

    Based on the tone of .0, I'd say you are definitely losing your
    mind.  However, that is probably not the problem.  I suggest you
    spend some additional effort on talking to your spouse about a hard
    drive.  They aren't very expensive any more, and you'll enjoy your
    computer a lot more when you have one.
    
    When I found that I needed a hard drive, I gave away the Apple II
    (will be a tax-deductable contribution if I play my cards right)
    and bought an A2000 with 3MB memory and a 40MB hard disk.  I then
    took over one whole room of the house to hold it and my other toys.
    My spouse put up with this because she has friends whose husbands'
    hobbies take them out of the home a lot: golf, fishing, hunting,
    attending football games, etc.  Computers is like stamp collecting
    or amature radio---it keeps you home.
    
    Tell her if you can't have a hard disk for Christmas, you'd like
    some golf clubs and fishing tackle.  Maybe she'll see the light.
    If not, contact me via e-mail and I'll see if I can get my wife
    to call her.
        John Sauter

    
    Naw, there's no way that the system would ask for the SYS: disk
    (in your case, DF0:) just because you popped a new floppy into DF0:.
    I believe that all the system stuff needed to recognize a floppy
    is in the trackdisk.device, which is in ROM, or KickStart on the
    A1000.
    
    Also, the WB disk shouldn't have accessed just by popping your empty
    'Awesome_Demos' disk into DF1:.
    
    I'd say that your 'Awesome_Demos' disk wasn't properly validated.
    Try repeating the problem with a freshly formatted disk.  If it
    repeats, I'd say you have a bad drive or something.
                     
    By the way, if your twin floppys are improperly cabled (this is
    highly possible given that various brand drives that CBM uses can
    have connectors on upside down with respect to each other) you can
    see symptoms like you described.
    
    Ed.
    

    re:.0
    
    write enable the floppy that causes the disk-validator to be loaded
    when you put it in the drive.  Let the validator do it's thing to
    it.  Wait a bit to make sure it is done.  Then write protect the
    disk.  I've seen the same problem with disks that were 100% full,
    the disk-validator needed to make sure it liked the disk.  After
    it gives the disk the "stamp of approval" then the system won't
    need to load the validator.
    
    -dave

    re: -1
    
    I seem to remember reading something about the Disk Validator being
    invoked *each time* a completely full disk is inserted; had something
    to do with the fact that on an almost-completely-full disk, there
    is no room to write the blocks-in-use bitmap, which DOS tries to
    read when you insert the disk.  If it doesn't find a good map, the
    validator is invoked to rebuild it (which of course fails if there's
    still not enough space on the disk).  This is from memory and the
    details may not be accurate, but I'm pretty sure the general idea
    is correct.
    
    But you were using an empty disk...perhaps it just needed to be
    validated once?
    

    I think that the bug is far more subtle than just a problem with
    the disk validator.  I can take a disk, produce a perfect copy using
    Marauder, and make 1.3 go OTL.
    
    It seems that 1.3can't deal with two IDENTICAL disks - in this case,
    that means the name and the incarnation number (derived from the
    date).
    
    If you put both disks into your Amiga (assuming you have at least
    2 drives), 1.3 can't figure out which disk is where and deletes
    both disks from its tables.  The icons disappear from the WB screen.
    If you then remove one (or both) of the disks and do a directory (or
    similar), 1.3 will ask for the disk to be inserted (even if one
    copy is still in a drive (any drive)).  Putting disks in, both disks
    in, doing DISKCHANGE, or having both disks out produces no change
    in the effect.  The disk (and its duplicate) are now unreadable
    until your next reboot.
    
    How, you may ask, did I discover this?  You see, I made a copy (using
    Marauder) of my 1.3 Workbench distribution so I could hack on the
    copy and happened to have needed to have both disks in the system
    at once because I accidentally deleted a critical file.  Can you
    imagine the confusion this can cause when the problem is confusion
    over the Workbench disk?!?!?  Sigh...
    
    BobW
    Now, children, don't try this at home because its very dangerous.

    re:.5  it's not a bug, it's a feature.
    
    Your disks are tracked by the OS by the disk ID.  It hates to find
    two disks with the same, identical, ID.  It's a small price to pay
    for an OS that will ask for the disk it needs, instead of blindly
    writing to whatever disk is in the drive.
    
    -Dave

    See note 2006 for the same problem with a public domain program,
    TurboBackup.

Re: .5

This "aspect" (that disks with identical internal disk ids are considered
to be the same disk) has been part of the Amiga's operating system since
day one.  It's not new to 1.3.


Re: .4

>    I seem to remember reading something about the Disk Validator being
>    invoked *each time* a completely full disk is inserted; had something
>    to do with the fact that on an almost-completely-full disk, there
>    is no room to write the blocks-in-use bitmap, which DOS tries to
>    read when you insert the disk.

One of my replies (lost somewhere in this conference) wrote up this
behavior.  The disk validation will succeed, but the system will still
not be able to write out the bitmap until you delete some space from
the disk (or write-enable it).


Re: .0

There are two disk validators.  One is memory resident all the time
and just makes sure the disk appears reasonable.  If the little validator
fails, the big validator is loaded from L: and performs be full validation
plus (if possible) repair (write out a valid bitmap).

I believe that the Fast File System's "big" validator is part of the file
system code itself.

    Hmm...  I thought I posted a reply in here before I went away for
    the holidays.  Snort!  Must've been too much 'nog' in my egg-nog
    (hic).  
    
    Anyway, I poked at this problem a little more.  I think it's the
    previously mentioned 'full-disk' problem.  The reason it got my
    AWSOME_GRAPHICS_DEMOS disk is that the disk was a copy of another
    full disk.  I was going to start with that, delete unwanted files,
    and then add new demos.  I changed my mind and decided to initialize
    a blank disk and fill it up instead, but I hadn't put labels on
    either of the new disks... and what with all the Christmas spirits
    I'd been having, I mistook the non-empty disk for the empty one.
    Well, just what I deserve for drinking behind the wheel of my A500.
    ;-)
    
    
    					John 
    
    BTW, John... It worked.   I mentioned golf clubs and fishing tackle
    instead of a hard disk and my wife came around nearly instantly.
     Who knows?  If the budget can swing it, I might even talk her into
    a 2000?   
    
    

    re: .9
    
    Glad things worked out for you.
        John Sauter