[Info-vax] Price/Value of Used VAX 4000/700

[Johnny Billquist]

On 2011-10-24 20.44, John Wallace wrote:
> On Oct 24, 10:47 am, Johnny Billquist<b... at softjar.se>  wrote:
>> On 2011-10-24 08.23, Hans Vlems wrote:
>>
>>
>>
>>> On Oct 23, 6:18 pm, "Richard B. Gilbert"<rgilber... at comcast.net>
>>> wrote:
>>>> On 10/22/2011 11:28 PM, Howard S Shubs wrote:
>>
>>>>> In article<pan.2011.10.21.16.57.15.932... at swbell.net>,
>>>>>     Jim Hopkins<bwana... at swbell.net>      wrote:
>>
>>>>>> Thanks to all for the input. I think I will probably just offer to
>>>>>> haul it off after initalizing the disks in their presence to clear
>>>>>> the data, which is probably their main concern. I'll likely have some
>>>>>> hardware questions for the group if and when I get it.
>>
>>>>> We'll be here.
>>
>>>> Initializing the disk DOES NOT remove any data on the disk.  It does
>>>> make it a little difficult to access the data but it's by no means
>>>> impossible.  What it does do is remove the pointers to the data.
>>
>>>> If you want to remove your data, write all ones to every byte in every
>>>> block.  Then right all zeros to every byte in every block.
>>
>>>> This WILL NOT stop the CIA or the FBI from recovering data from the
>>>> disk.  If you need to need to prevent access to the data on that level,
>>>> Google for "DOD Erase".
>>
>>>> Or bake the disk at 500 degrees Fahrenheit for two hours.  This should
>>>> guarantee that the magnetic domains have all been randomized.
>>
>>>> Better yet, don't commit anything secret to disk or tape.
>>
>>> Baking a disk at 500 degrees Fahrenheit very likely demagnetizes the
>>> magnetic coating.
>>> IIRC magnetic materials have an attribute called the Curie temperature
>>> above which the
>>> material (permanently?) looses its magnetic properties. If so, then
>>> your method will both
>>> destroy the data on the disk as well as its usability!
>>> Writing all ones, or all zeroes or any random bit pattern to all
>>> blocks on a disk is not sufficient
>>> then?
>>
>> Depending on your budget, and how desperately you want the bits, you can
>> actually read through several overwrites.
>> But this is not something normal people ever have access to do. That's
>> where the "CIA or FBI" comment comes in.
>>
>>          Johnny
>
> People don't really still believe these tales about magnetic remanence
> and DoD erase patterns do they? Have a look at how the signal
> processing works in any recent disk drive, and then come back and tell
> me how it can still credibly be done.
>
> It may well have been plausible in the days of the 2.5megabyte RK05,
> when heads could be commanded to go off-track (and naturally went
> gradually off-track anyway over time), and there was enough magnetic
> remanence to make the stories plausible (if not always demonstrable).
>
> But some years later, we're now in the era of embedded track servos,
> unbelievable track densities, and things like partial read maximum
> likelihood data recovery, where it's a minor miracle the data is
> readable at all, and the chances of guessing what was in any block
> prior to the current overwrite are zero.

So, tell me. Do you think the magnetic field will be equally strong if 
you write a one over a one, or a one over a zero?

Remember, I'm not talking about reading out the end result with the 
normal read/write head, but doing a read of the actual strength of the 
magnetization.

It's not about trying to read data off track center. You have never been 
able to read previous generation of data from a disk with the normal 
heads. They are not designed for that purpose, but rather the opposite. 
You do not want to accidentally read the wrong data, so you design it 
for the maximum safe thresholds.

	Johnny