Moe Trin, from an Executives Office on the 120th floor, proclaims to our
ignorance:
> In the Usenet newsgroup alt.os.linux.suse, in article
> <dcc6ob$fg8$1@godfrey.mcc.ac.uk>, Roy Schestowitz wrote:
>
>>Interestingly enough, the issue of formatting came up yesterday in another
>>newsgroup that I participate in.
>
> The concept of formatting erasing data is an Urban Legend that goes back
> to DOS. People whose only exposure to DOS was "MS-DOS 7" (the so-called
> DOS prompt in windoze95) know this as a fact - despite changes dating
> from MS-DOS 5, never mind the way it was with PC-DOS 2.0. Those who never
> even saw a DOS prompt (win98 and later) are even more positive this is
> the case. With _floppy_ drives, the 'FORMAT A:' command _did_ wipe
> the disk, because the original procedure had a media write test - the
> media being so unreliable. However, the largest floppy every produced was
> the "2.88 Megabyte" 3.5 inch (90 mm) floppy, and even at that density, a
> format should only take about 90 seconds. As media became more reliable,
> MS-DOS 5.0 changed the default mode of floppy formatting to the fast
> format, where the command only did a read test, saved a copy of track 0
> and 1 (the "unformat data"), and then overwrote the outer track (clearing
> the DOS Boot Record, FAT tables, and root directory). There was also a /Q
> option that eliminated the read test - only saving the unformat data, and
> clearing the first track. The /U option did the old style "wipe the disk"
> style format, unless combined with the /Q option (when combined, it ONLY
> overwrote the outer track - a few seconds at most).
>
> Hard drives are larger, and take proportionally longer to format in the
> same way as the original floppy method. Originally, hard disks spun at
> 3600 RPM (10 or 12 times faster than a floppy, but the limitation is
> actually the bandwidth of the controller. I vaguely recall the low-level
> formatting of a Seagate ST419 (15 MB) taking about ten minutes. I also
> remember using Gibson Research "Spin-write" to change the interleave on
> a ST-251 (42 Meg) on an AT (286-8), and the process taking well over an
> hour.
>
> Consequently, the 'low level format' was done just once. Then you used
> FDISK to partition the drive, and 'FORMAT' to create the index tables
> of where stuff was on each partition. The various manuals used to warn
> that you would loose all data if you low level formatted (true), but
> also if you repartitioned or formatted the hard drive. Likewise, you
> lost the data when you 'DEL' or 'ERASE' a file. Does anyone remember
> Peter Norton (before the company got borged) and what he "invented"?
> Back in the 1980s, when I supported DOS, the Norton Utilities were
> used quite often to recover from a user screw-up.
>
>>It appears as if many people begin to worry about safety and security
>>of their data these days; much more than they did in the past.
>
> The theft rate for computers has been fairly high - especially for the
> smaller laptops and the like. They are easy for a thief to grab and hide,
> and can quickly be converted to cash. Additionally, people are putting a
> lot more information of value on them, and usually don't take the time to
> see that the data is secure - encrypted in some manner. Most people who
> deal in stolen computers will erase the data (or so they think) on the
> computer before selling it to reduce the chance that the system will be
> identified as probable stolen property. Especially with the larger drives
> of today, the thief, or receiver of stolen goods won't spend the time
> snooping through the disk looking for interesting stuff, unless the
> original owner was targeted for some reason, or had made things obvious.
>
>>The IBM fingerprint authentication (IBM T42 running Linux) was the point
>>of focus in that discussion. Firmware avoids access from another computer,
>>fingerprint authentication (maybe key) enables restricted access to data
>>on the original machine/housing.
>
> Restricting access from another computer is easy, but where the cheese
> gets binding is when someone has physical access to the hardware. I don't
> care what software you are running, if someone has access to the hardware,
> it's _ALMOST_ always open to everything. There are some mechanisms such
> as disk encryption that will go a long way in protecting things in a case
> like this, but physical security (locks on doors, windows, and the
> computer case, alarms, and/or guards) is the bottom line.
>
> Old guy
I love reading your analysis, but I'll warn you now that a discussion which
goes back to the 'format /q' days takes me back to my early teens, at which
stage I did not have any fine-level understanding of computer. In fact, the
granularity of my knowledge at the time was laughable.
I have always believed that when formatting a floppy disk, going through the
sectors involved stuffing them with arbitrary bits or just zeros. I could
imagine that quicker forms of format were not a magic solution, but all I
had on floppies at the time were games, though rarely I had some shareware,
lightweight programs.
As for hard-drives, formatting time always appeared miraculously short,
which ought to imply that the process was rather shallow. I did not format
many hard-drives in my life though. I can recall the duration of one hour
or so, but I might be wrong. Maybe my memory betrays me.
I still remember the days of undelete.exe and tools like Norton Disk Doctor
(along with other utilities from the prolific Norton). They were a true
evidence for the fact that only file allocation tables were affected in the
interest of speed gains. Sadly enough, there was never a way (a /common/
tool) of scrambling all unused sectors on a drive.
You must be a big storage expert as your discussion teaches me plenty while
there is little that I can contribute without eroding the quality of your
words. Because of the depth of this ~4-day old thread, your words have
gotten undervalued although they beg reaching the attention of a wider,
well-justified audience.
All the best,
Roy
--
Roy S. Schestowitz
http://Schestowitz.com
|
|
