P-list is a SCSIism, but I'll agree that every drive will have one
like it.
> moved to spare cylinders so the head seek causes the slower access
> time, similar to the random seek times that HD Tune scatter-plots. The
> values I see are somewhat more than 50mS on old, slow drives, compares
> to >100mS for weak areas. I've used the information only once, to
> partition around a large bad area on an otherwise decent Hitachi 7K60.
If your application is so real-time critical you ought to insist on
using a drive that doesn't have marginal/remapped areas to begin with.
Some filesystem formatting utilities will accept lists of bad blocks,
so this doesn't necessarily have to be handled at the partition level.
> HDD Scan reports the response time for each individual slow block,
> apparently requested individually to avoid the uncertainty of
> buffering. Somehow it also can read S.M.A.R.T data from a drive on a
> USB cable.
Depends on the USB->[S]ATA interface. I have seen interface
adapters that don't handle the SMART command subset and some that
will.
> > These days, you can't eve trust software for which the source code
> > is
> > available unless you can afford to do a full security audit. ...
> > Uncle Steve
>
> I don't. I maintain an updated master system on a good new drive in a
> CD bay caddy and copy it to the old drives I use to browse the net and
> test software. I doubt that malware on the 'sandbox' drive can migrate
> backwards to a bootable DVD backup in a read-only DVD-ROM drive. XP
> and 2005-vintage laptops probably aren't the prime targets of hackers
> anyway.
Yeah, but how do you verify the integrity of the software in its
original form? I am, of course, referring to factory malware. And
there is the bug/feature dichotomy that doesn't always lend itself to
trivial categorization. I have some HTTP proxy software that for some
reason will not cache certain URLs, even if the program is instructed
to "violate" the HTTP standard. The consequence is that certain
sites that gather a whole lot of profiling information about their
visitors cannot be foiled.
www.drudgereport.com is one such site. Is
it coincidence that that site is structured in such a way as to foil
the aggressive forced-caching of the proxy server, or are the authors
of the software in bed with the NSA? Impossible to say with any
meaningful degree of certainty. I've not named the package as I have
not yet done sufficient analysis of the code to make strong
accusations, nevertheless the observed behavior is extremely
suspicious.
So forget malware as as layperson might conceive of it. The range and
scope of potential attacks on your computer, software, and data are
huge. And without the source code, you have no hope whatsoever of
making any determinations of security or completeness. That goes for
your applications, the operating system, and the hardware. For the
hardware, you need the VHDL code for the silicon in each chip on a
bus, as well as assurance that the chip fab is in fact laying down
the gates specified by the VHDL and nothing more.
You, and everyone else have zero assurance that you aren't running
computers that are wide-open to arbitrary penetration by the big TLAs
and their partners in organized crime. Just because you might say it
is paranoid to have those concerns is no reason to assume anything
about the security of your computer, phone, television, radios,
stereos, digital watches, or hearing-aids.
The usual counterargument is for someone to say that they aren't doing
anything wrong, so they don't have anything to fear. Well, I haven't
done anything wrong, but the government (in part) has sought to run my
life for their shitty political purposes. And they have been ruinous
to my data and electronic appliances in the process, in part because
my socio-economic isolation is more convenient to their purposes.
It's called Plutocracy, and it's not just for "Communist" countries.