Conference hydra::amiga_v1

    
    The inhibit option is also in the CBM command, it will turn the
    respective drive "busy" so that you can insert and format a disk
    that otherwise might drive AmigaDOS crazy.
    
    Regards,
    
    <CB>

Re: .1

	Oh, thanks.  I 'spect I was using the CBM format command, since it's
not one of the ones ARP replaces.  The option isn't listed in the 1.3 enhancer
manual, though.  I was hoping it did something like inhibiting the future use
of blocks that had been found bad.  With what may be a flakey hard disk on
my hands, some way of doing this would be nice.

Mark

    
    Bad block mapping should be handled by the driver/handler/device/younameit
    and not at the DOS level. There has been a rather longwinded discussion
    on UseNet about this some time ago, and it was decided that this
    was the way to go.
    
    I recently connected a 2nd RD53 to my A2090 and happily ignored
    Prep's asking me for bad blocks and formatted the beast with the
    normal DOS format. After that I had 41 blocks in use on an empty
    HD. I filled it, mutilated it, and no errors.
    Hmm, seems like the 2090 device does a hidden bad block mapping
    after all. :-)
    
    Regards,
    <CB>
    

But will it do bad blocking on a SCSI device?

John

    Re: .4
    
    I don't have the slightest idea...
    Why don't ya try simply try it???
    And as far as I know, Prep will issue a low level format on the
    SCSI device, and usually bad block handling and remapping is done
    by the SCSI unit itself.
    
    And just don't ask me what I'm doing THAT late in the office...
    
    Regards,
    
    <CB>
    

    re: -1
    
    Arrrghh, even my English starts failing me, quick gimme that vacation,
    will ya? :-)
    
    <CB>

	Here's the situation:  Some months ago (when 1.3 was released), I ran
Prep on my ST-251-1 connected through an A2090, and then formatted it with
the format command in the system draw.  Last week, I had problems - some
commands just weren't there anymore, and I was getting a requester stating
that the disk had a read/write error occasionally.  I reformatted using the
format command in the system draw (of my hard disk maintanance floppy), and
restored the disk from my backups.  I did NOT run Prep.  During the format,
there were a few cylinders that took longer to format than the others.

	Should I have run Prep?  Would it have done any good?  The Hard disk
driver is the one I got with the A2090 well over a year ago (before there was
an A2090A).  Will that take care of bad blocks?  I thought, at the time, that
that wasn't yet implemented, though several people were requesting it.

Thanks for any info,
Mark

    
    Re: .7
    
    As I stated before, Prep SHOULD do you a lot of good when you run
    it on a SCSI unit, since these suckers are supposed to do their
    own bad block handling. If it's a ST-506 device, it will only help
    you if know where bad blocks are. I remember reading something about
    a PD program that does bad block detection on Floppies and HD's
    here or on UseNet, anybody has it?
    My experience (not knowledge) is that if format doesn't report a
    read/write error during format (but takes a little longer at some
    cylinders), some sort of bad block mapping is done. Don't ask me
    which, it just happens to work for me. 
    Ask Steve Beats from CBM on UseNet if you're interested in any intimate
    details. ;-)
    
    <CB>
    

Is there some inherent limit in the A2090a on number of partitions or size
of drive?

I am trying to add a Quantum 105S to my system but can't get it partitioned
the way I want.

It seems that whenever the size of a single, or total of all partitions
exceeds about 50 MB, I get trouble.  I can mount all partitions regardless
of size/number (MOUNT doesn't complain). Only the ones under the ~50MB limit
show up with INFO. I can formet them and use them no problem. If I mount
and attempt to format the ones outside this "limit", Format complains:

		format failed
		can't find handler

Anyone know what's going on here?

John

P.S.  This sucker peaks out at 250 seek/reads/second and 655 Kbytes/second
	with diskperfa!

    Re: .9
    
    Nah, no, njet, nada, nein.
    
    I currently have one RD53 (70MB, ST-506) on my 2090 divided into
    four equal sized partitions and another RD53 as one large partition.
    
    Everything works just fine for me.
    
    I formatted the large partition just two weeks ago with no problems.
    I also had a 80MB SCSI unit once attached to it and formatted as
    a large partition without a hitch.
    
    There had been a limitation at about 52MB for Old Filesystem
    partitions, but the diskperf numbers you give clearly state that
    you use FFS.
    Under FFS the max. partition size is 2.5 GigaBytes.
    
    Beats me. (pun intended ;-) ) 
                                                     
    <CB>

This drive uses "zoned bit recording" so it reports the number of sectors
per track as 0 (zero).

Would the 2090a be issuing a MODE_SENSE command to get this info and be
getting messed up?

Is there any technical docs on the 2090a - like how to send a SCSI command
directly??

John

    
    Re: .11
    
    Since my first reply was eaten by the flakey net or Bombe::'s
    connection to it, here I go again:
    
    Maybe it's Mode_Sense, but I don't think so.
    
    The next release (whenever it may surface) of the 2090(a) drive
    will/should support the hardblocks and SCSI-direct protocolls.
    These would help you alot.
    
    In the meantime, get a copy of the A500/2000 Technical Reference
    Manual.
    
    Regards,
    
    <CB>

>    In the meantime, get a copy of the A500/2000 Technical Reference
>    Manual.

Where can I get this?

John

    Re: .13
    
    The TechRefMan is only available from Commodore, as far as I know...
    
    <CB>
    

    Here is the way to order Amiga technical documentation, taken from
    the monthly Usenet "Introduction to comp.sys.amiga.tech" posting.
    
    
From: cbmvax!lauren (Lauren Brown CATS)
Subject: Re: Tech Ref. Docs
 
This list will be growing, as we make the 1.3 Native Developer Update
available, along with some other stuff we're planning.
 
    Commodore-Amiga Technical Information
    =====================================
              
Amiga 1000 Schematics and Expansion Specifications
  Full A1000 schematics, timing diagrams, PAL equations, and 
  documentation of auto-config process  
  
IFF Manual and Disk
  Full IFF documentation and source listings, including new IFF
  form chunk ANIM. IFF disk includes source code, object files, 
  executable programs and documentation.
 
Fall 1986 Developers Conference Notes
  Contains the diagrams, outlines and additional notes pertaining
  to each conference speakers' topic. Some additional information
  has been included (e.g., 8520 specifications).
 
AmigaMail
  Bimonthly technical newsletter produced by CATS. Articles range from
  programming standards through programming for compatibility to 
  advance technical information on new products (hardware and software).
 
Cost EACH for any of the above:  U.S. $20 (in Canada $22.50, $25 elsewhere)
 
A500/A2000 Technical Reference Manual
  275 page reference manual that describes the technical
  features of the A500 and A2000, and the features that differ
  from the A1000. Included are: System block diagrams, Amiga
  Expansion, Designing hardware for the Amiga Expansion Arhictecture,
  Driver documentation, Software for Amiga Expansion, BIOS entry
  points, Clock/Calendar Registers, plus much more. Also included
  are schematics.
 
Cost each:  U.S. $40 (in Canada $42.50, $45 elsewhere)
      
        TO ORDER
        ========
 
Send check or money order payable in U.S. funds to:
 
        CATS-Orders
        1200 Wilson Drive
        West Chester, PA 19382