Conference akocoa::decpc_portables

    SNLV01::LEE
    
    Just repair the notebooks then make sure you charge the customer 
    a hefty fee for the user induced problem.
    
    /Bill (only half kidding)

   [This is not intended as a flame.]

:    Just repair the notebooks then make sure you charge the customer 
:    a hefty fee...

   ...for a notebook design problem?

   --

   For an operation this simple, we should not need to refer folks to the
   manual for subtleties -- the hardware should do what is expected...
   Whether or not we told the user `not to do this', merely having this
   sort of less-than-stellar system behaviour does not reflect well on
   DIGITAL PCs, nor on DIGITAL designs in general.  (Yes, I know, these
   systems are buy-outs.  DIGITAL's name is still on the box...)

   There are more than a few ways to scramble your data or damage your
   hardware with a PC-class system.   Lets not add any more ways, please.

    
    	Well, for now, the customer's got a machine that happens to have
    	this "design problem" thrown in at no extra charge.
    
    	So the only thing that can be done for the base noter is to sup-
    	ply a method of data retrieval after this has been done, and a
    	bit of education directed towards the customer so that [s]he is
    	aware that it shouldn't be done in the future.
    
    	Why'd the customer buy this particular machine if hot-swapping
    	was such a necessity?  How long does it take to suspend the mach-
    	ine and swap the drives correctly?  Or if both drives are needed
    	at the same time, why not connect the floppy to an unused port
    	and leave the CD-ROM drive in the bay?
    
    	I have a non-Digital laptop, and it also recommends against hot-
    	swapping.  I know this because I RTFM.  And I also know that I
    	can hook up the floppy drive to an unused port, again because I
    	RTFM.
    

    re .2
    
    And what does it say for Microsoft that allows you to POWER OFF your
    system without first dumping the cache on almost any system known to
    man today. Should MIcrosoft have some magic way to prevent you from
    pulling the plug on your system or to keep you from shutting off the
    power. But you say, your system was just sitting at the desktop and no
    applications were running, No matter, NT still has not dumped the last
    disk updates and changes until you say Goodbye... Wouldn't you think
    that such a responsible company as them would find some way to save
    your data or disk...
    
    Removing the CD or Floppy drive on a portable is the same thing as
    going to your desktop and ripping out the IDE bus devices and then
    wondering if the spikes caused any damage to your equipment... 
    It is just a little less obvious because you don't have to take off the
    cover.  Hot-swapping some things are allowed, batteries qualify. But
    removing all batteries and not plugging in to AC will definitely cause
    a loss of operation.. Sorry only makes sense to me..   It has
    nothing to do with a buy-out or DEC designed or anything else but the
    laws of physics and poor understanding of how to operate machinery. You
    don't see many people trying to change a tire while driving down the
    highway at 65 miles an hour any more do you.. 
    
    Just my $.02
    Vin

   Microsoft has some serious hardware integration problems, and some
   serious file system coherence problems on `unexpected shutdowns'.

   If Microsoft software chooses to scrozzle it's disk, that's
   Microsoft's problem.

   My comment is that DIGITAL hardware doesn't fry its brains.

   As for specific comments, I understand there are some cases where
   disconnections can cause problems.

   I know this won't fly, but I'll suggest it anyway: software
   interlocks for the hardware.  (I also understand that Macintosh
   has handled things like disk ejection and the power-off button
   far better than Microsoft ever has.)

:    Removing the CD or Floppy drive on a portable is the same thing as
:    going to your desktop and ripping out the IDE bus devices and then
:    wondering if the spikes caused any damage to your equipment... 

   Not quite -- this stuff is documented, designed, and intended to
   be user-removed and swapped out.  We just depend on the user to
   synchronize this activity with the operating system, with no sort
   of interlock to prevent problems.  The rest of your point -- about
   removing widgets on the fly -- is otherwise quite valid.

   I hear a strong concensus of `the user screwed up!', and not of
   `how can we keep the user from screwing up?'...  Or "how can we
   avoid scrozzling the hardware if the user screws up?"  It's the
   latter cases that improve the product...

    I am told that with enough money and buffers, it is possible to buffer
    all of the connections and allow Hot Swapping of everything. I woul
    dlove it but they tell me it is not possible in a system costing under
    $2000.  Look for the full featured hot swapping on hte DEC higher
    priced models (Ultra 2 and future products of that category).
    
    We try very hard to make the system 'user proof' but there are just
    some things that can not be accomplished in the parameters of the
    product.. It is not because we did not complain, whine and scream, it
    just was not possible to do on the value line products and still
    produce a product..
    
    Yes caveats adn use beware signs are not nice but it is all we can do.
    If the user actually does read the little notices that we stick in the
    systems, they can probably enjoy their systems very much. It is a neat
    little box and FAST.... I use mine for daily desktop and travel use and
    believe me, I wish it could Hot Swap everything but alas...
    
    Vin

:    We try very hard to make the system 'user proof' but there are just
:    some things that can not be accomplished in the parameters of the
:    product.. It is not because we did not complain, whine and scream, it
:    just was not possible to do on the value line products and still
:    produce a product..

   Have you looked at the "soft power" buttons available on various
   systems such as Mac (we have a start on this with the VP 500), and
   considered some sort of interlock that forces the required software
   state... Or an interlock that prohibits the device removal when not
   in the correct state?

    re: .7 by XDELTA::HOFFMAN
    
    >> Have you looked at the "soft power" buttons available on various
    >> systems such as Mac
    
    Just fanning the flames, but are you suggesting that Macs have the
    capability to allow hot swaps of SCSI devices?  Else, what's the big
    deal with soft power switches and what problem does that solve along
    this topic's vein?
    
    >> some sort of interlock that forces the required software state... Or
    >> an interlock that prohibits the device removal when not in the correct
    >> state?
    
    Sounds electro-mechanical.  It'll be heavy, expensive, or suck up a lot
    of power.  Just a pessimist's hunch.
    
    /Bill

    
:    >> Have you looked at the "soft power" buttons available on various
:    >> systems such as Mac
:    
:    Just fanning the flames, but are you suggesting that Macs have the
:    capability to allow hot swaps of SCSI devices?  Else, what's the big
:    deal with soft power switches and what problem does that solve along
:    this topic's vein?

   No, I'm refering to the software control over the hardware that one
   has on a Mac -- there's no "eject" button (normally) used to eject
   the floppy from the drive, one asks the software to do it. This has
   some very obvious benefits around disk data caching...
   
   One also has a software-controlled power button -- it (usually) does
   the "right thing" [tm] when pressed.  (The HiNote VP 500 certainly
   has a good start here...)

   As for hot-swapping of SCSI widgets, no.  (Though if that can be
   done -- and various SCSI converters can quiesce the bus and allow
   this -- it's certainly going to be a useful feature.)

   Again, I'm talking about dealing with explicitly removable storage
   widgets in an integrated (and non-traumatic) fashion...

:    >> some sort of interlock that forces the required software state... Or
:    >> an interlock that prohibits the device removal when not in the correct
:    >> state?
:    
:    Sounds electro-mechanical.  It'll be heavy, expensive, or suck up a lot
:    of power.  Just a pessimist's hunch.

   Possibly.  But there are electronic sequences, sometimes involving
   a capacitor and/or pins that disconnect in a particular sequence.
   (Regardless of the implementation, my intent would be to allow the
   user to remove the device only at the correct time, or to allow the
   user to remove the device at any time without untoward effects...)