Arno said:
Actually verify on write deos help massively, as it identifies whether
a write was good or not.
For certain values of "massive" perhaps.
Doing it in software is not the same. If
the drive does it, it can use higher standards for successful verifies
than for standard reads.
If the drive has the hardware to do this then what prevents the software
from using it?
You are correct that it is not the only thing
that guarantees long-term readability. But chemical stability, once
assured, is not the main reliability problem of optical media.
"Once assured" maybe, but who says that it's assured?
There is a new technology developed for dealing with this, it is called a
"rag".
And why do they go bad? Because your "assured chemistry" is no such thing.
"Vibration"? Vibration removes data from optical media? I've never heard
that one before.
For example I had one MOD that
had > 900 reallocated secotrs, because dust got into the cartridge.
Well, now, that's one of the problems with cartridges.
Write only these would all have been potentially unreadable sectors,
possibly after some time (since the initial write could have been
borderline, but a slight degradation could have made them unreadable).
And what would cause such degradation?
But the drive identified them on write and made sure that all data
was written well. And at the end that I was informed that tha last
write was not successful.
Geez, Nero would have done that for 60 bucks.
I happen to know that this was dust, since
after cleaning and reformatting the same MOD has worked well and
reliably without significant additional defect sectors for several
yeras now. True, MOD is not quite the same chemistry as DVD-RAM,
MOD does not rely on a chemical change, it relies on a change in the
magnetic properties of a material with temperature. It's no more dependent
on chemistry than a tape or magnetic disk is.
but long-term stability is not really an issue with the chemical
structure if the disk was manufactured according to standard (a
real issue with El Cheapo DVD-RWs!).
To _what_ standard? Does the DVD-RAM standard mandate a given chemistry?
There is A;sp a second thing: MOD/DVD-RAM are factory certified and
have an initial defect list, were the manufacturer found weak
spots with _very_ sensitive equipment. No such certification is
done for DVD-RW.
Well, that's nice, but it still doesn't make the chemistry stable.
As I said the chemistry is not the issue with DVD-RAM.
Chemistry _is_ the issue. You can have absolutely perfect write and if the
chemistry decides to go to Hell a year down the road you lose the data.
This has been the major problem with DVDs for archival storage right along.
There are chemistries that are supposed to be reliable but no guarantees
that any given media use those chemistries.
That may be a bit naive. Who has the time to polish their disks?
One only polishes the disk if the disk is damaged. This is an infrequent
occurrence.
Wipe, blow, or brush it off. Geez.
If they cause a problem, you clean them off of course.
What about them? Unless you roll your chair over it drops generally aren't
an issue.
not correctly
aclingned in closing drive, ...
If the drive closes at all you open it and reinsert the disk.
It amazes me that you trust the one thing that you can't inspect, the
chemistry, but worry about all this stuff that doesn't usually cause
problems in the real world and is generally easily correctable if it does.
Not an issue. Jous clean it off with soap and water. The data is still
there.
After you take the cartridge apart.
Yes, and the stated reliability drops dramatically for the formats
without cartridge, or once a disk has been removed from cartridge.
You trust "the stated reliability"? My respect for you has just dropped a
great deal. If it's mission critical data never trust anybody.
1.) Yes.
2.) Backups and archiveing would have been far more work and worry
Uh, copying it to MO _is_ "backups and archiving".
Floppies: About half of my older 3.5" floppies were unreadable
when I backed them up 2-3 years after they were last used.
There is something wrong with your hardware, Arno. You've been having all
of these problems, hard disks crashing right and left, diskettes unreadable
after 3 years, etc. You really need to get to the bottom of it instead of
just accepting that "that's the way it is".
CD-R: First loss after 2 week on a disk that verified fine.
Numerous other losses or parial readabilities. And
a bitch to use.
I don't see cranking up Nero or Roxio and dragging and dropping as being "a
bitch".
CD-RW: Disk that verified fine in the writing drive, but was
unreadable in others. Also see CD-R.
Only one, or multiples? If multiple you've got a rum drive. MO can suffer
the same problem.
HDDs: Two deathstars and two fried Fujitsus (the second my
mistake). The deathstars with warning signs to late to
get all data off.
So? If you have to "get the data off" then you aren't serious about your
data.
Cheap tape (Quic-80): A nightmare. Compares that did not compare.
Tapes that were unreadable after a few weeks.
Well, yeah, that's QIC. Totally worthless for any purpose. Tapes aren't
heavy enough to use for paperweights and aren't strong enough to prop up a
table leg and the drives are too light to work as a doorstop.
I have no experience with professional storage tape, but I
expect it comes on par with MOD for reliability, has not as
long data lifetime (say >10 years) and is far cheaper per byte
and far more expensive per installation.
DLT has at least a 30 year storage life. Not accelerated aging, on the
shelf. For equivalent capacity the cost is about the same as MO. The
thing is a 20 gig enterprise-grade DLT drive that cost $5K new is now
obsolete surplus that you get on ebay for 50 bucks in good working order.
The tapes cost about the same as MO cartridges.
Yes, I would have definitely lost data. Unless I had installed a
rigorous "3-copies and check each every year" regiment for my
backups. In contrast MOD is a "write and forget" technology, that
does not require any drivers or special softwware and will reliably
keep your data/family photos/stuff you care about safe for decades.
How would you like a nice bridge? I've recently come into the possession of
one. If that's not to your taste I also have a ski resort in Iowa and some
nice waterfront property in Louisiana. You find it trustworthy only
because it hasn't bitten you yet. Never, _ever_ trust one copy of
_anything_.
(Most vendors say > 50 years, a Phillips engineer told me probably
shaky...).
They claim that for DVD-R too. Why do you believe the one and not the
other? Accelerated aging models _are_ shaky until they've been correlated
with reality. Epson got bitten a while back with their archival
inks--seems that there was some common set of conditions that they didn't
allow for in their accelerated aging, and their prints were turning orange
in a matter of weeks for some users.
Interesting side note: In Japan MOD is all the rage. Seems people
there have more of a long-term outlook on storing data.
Or they're gadget junkies. The fact that something is the rage in Japan
doesn't make it useful. Remember tamaguchis? Or Aibos? Or totoros?
Second side note: In medical imagery, the law here requires the
images to be kept available for 20 years. It seems all the
scanners use MODs for this.
Which is fine, but kind of beside the point when you were defending DVD-RAM
earlier.
Personally the last time I lost data was about 20 years ago. Learned to use
RAID and tape and redundant servers and avoid a technological monoculture
and have little problem. Anything that I really care about is on at least
two servers, sometimes three, in separate locations and using different
operating systems. Lot of hassle you say? Nahh, it's all automated.
