late experiences with fat32

I have been having some of the same old problems lately with my computer
that I have had off and on for a few years now. Well it's old. I'd say the
ram is going bad. They're DIMMs. but my computer decides it will restart
itself. over and over. Well I lost a file. And OE is not resetting newsgroup
compression and such like it should. And firefox's settings are gone.

So is this the vulnerabilities of fat32? I remember fat16 and it was
terrible. I converted to ntfs. Maybe that will be a better option.

Maybe fat32 does have some weaknesses. Am I experiencing them here?
Whew.

Bill

On 20 Oct 2015, "Bill Cunningham" wrote in
microsoft.public.windowsxp.general:

So is this the vulnerabilities of fat32? I remember fat16 and it
was terrible. I converted to ntfs. Maybe that will be a better
option.

Maybe fat32 does have some weaknesses. Am I experiencing them
here?

No. FAT has some weaknesses compared to NTFS, but randomly losing
data is not one of them. If anything, you're experiencing the
effects of a failing disk drive. NTFS might be able to better to
recover from disk errors than FAT, but if the hardware is going and
it will be gone soon no matter what file system you use.

Checking the error log, checking the S.M.A.R.T. statistics, and running
CHKDISK will give you more information about the integrity of the
drive.

Bad RAM can also cause data to be lost.

"Nil" wrote in message
...
On 20 Oct 2015, "Bill Cunningham" wrote in
microsoft.public.windowsxp.general:

So is this the vulnerabilities of fat32? I remember fat16 and it
was terrible. I converted to ntfs. Maybe that will be a better
option.

Maybe fat32 does have some weaknesses. Am I experiencing them
here?

No. FAT has some weaknesses compared to NTFS, but randomly losing
data is not one of them. If anything, you're experiencing the
effects of a failing disk drive. NTFS might be able to better to
recover from disk errors than FAT, but if the hardware is going and
it will be gone soon no matter what file system you use.

Checking the error log, checking the S.M.A.R.T. statistics, and running
CHKDISK will give you more information about the integrity of the
drive.

Bad RAM can also cause data to be lost.

I got the computer in 2004.

On 10/20/2015 12:49 PM, Bill Cunningham wrote:
I have been having some of the same old problems lately with my computer
that I have had off and on for a few years now. Well it's old. I'd say the
ram is going bad. They're DIMMs. but my computer decides it will restart
itself. over and over. Well I lost a file. And OE is not resetting newsgroup
compression and such like it should. And firefox's settings are gone.

So is this the vulnerabilities of fat32? I remember fat16 and it was
terrible. I converted to ntfs. Maybe that will be a better option.

Maybe fat32 does have some weaknesses. Am I experiencing them here?
Whew.

Bill





If you have bad RAM , then yes, you can lose data, you need to replace
it ASAP.


There is a possibility however that the contacts just need to be cleaned.

Clean and reseat then run memtest.


If it finds *any* errors no need to keep running the test...the RAM is bad.


As to Fat32 vs NTFS


I have experience here and can assure you NTFS has much better fault
tolerance characteristics.


I've seen fat systems be rendered useless by writing the data to a slew
of "chk" files

Bill Cunningham wrote:
I have been having some of the same old problems lately with my
computer that I have had off and on for a few years now. Well it's
old. I'd say the ram is going bad. They're DIMMs. but my computer
decides it will restart itself. over and over. Well I lost a file.
And OE is not resetting newsgroup compression and such like it
should. And firefox's settings are gone.
So is this the vulnerabilities of fat32? I remember fat16 and it
was terrible. I converted to ntfs. Maybe that will be a better option.

Maybe fat32 does have some weaknesses. Am I experiencing them here?
Whew.

First thing to do is test your memory with something like memtest86+. Bad memory
can cause all kinds of problems.

"Bob F" wrote in message
...
First thing to do is test your memory with something like memtest86+. Bad
memory can cause all kinds of problems.

It froze up while doing things. Test 5 had 223 errors and test 7 had 1
error. idk what that means.


Bill

Bill Cunningham wrote:
"Bob F" wrote in message
...
First thing to do is test your memory with something like memtest86+. Bad
memory can cause all kinds of problems.

It froze up while doing things. Test 5 had 223 errors and test 7 had 1
error. idk what that means.


Bill

If the test shows errors in the same locations
each time you run it, those are "stuck at" faults.
That means immediate replacement.

If the test shows errors in random locations, faults
like that may be cured by making adjustments in the BIOS.
The assumption is, a "speed" fault, which may be amenable
to tRAS adjustment, or voltage adjustments.

As an example, out of the blue my current machine started
throwing random errors. The VDimm was already at
the recommended voltage, and on a hunch, I adjusted
Vnb (Northbridge memory controller) by increasing the
voltage by one notch. The errors stopped.

If you have "stuck at" faults, you can test the memory
one DIMM at a time. That's to make sure you get a positive
ID of the bad DIMM. Remember to use antistatic precautions,
and turn off all power to the machine before adding
or removing DIMMs. You don't want any standby power
in the RAM slots, when pulling or inserting DIMMs. I
just unplug the computer to be sure.

For antistatic precautions, bring your body to the
same potential as the chassis, with an antistatic strap.
Static electricity works on "potential difference", and
it isn't important for you to be grounded (everything
doesn't have to be ground-references). Rather,
your body, the antistatic bag or box with the DIMM in
it, and the computer chassis should at be at the same
electrostatic potential, when transferring the sensitive
electronics. Static zaps, only jump between potential
differences.

(Preferred strap style - clip the alligator clip to a
screw on an I/O connector on the back of the machine.)

http://comingsoon.radioshack.com/ant...d/2762395.html

While electronics have some (small) amount of antistatic
protection (1kV or 2kV), you still need to treat the
stuff with respect. Because you cannot see when you
damage something, via bad handling. We had people at
work, who watched for bad handling techniques and
delivered lectures to staff. Our lab was antistatic
from top to bottom. But we avoided mandatory heel
straps (which I hate). That was reserved for rooms
with laser-based equipment. I would have bought a pair of
ESD shoes if they insisted on heel straps. The floor
and all bench surfaces were antistatic, and we wore
straps while working on stuff (soldering components
to prototypes). Even though the prototype equipment
wasn't going to customers, it was still treated
as if it was. Kind of a "training exercise" or
industry-best-practice for the staff. It's so if
staff went to the factory, they wouldn't
"embarrass" themselves by doing stupid stuff...

Paul

In message , philo
writes:
On 10/20/2015 12:49 PM, Bill Cunningham wrote:
I have been having some of the same old problems lately with my computer
that I have had off and on for a few years now. Well it's old. I'd say the
ram is going bad. They're DIMMs. but my computer decides it will restart
itself. over and over. Well I lost a file. And OE is not resetting newsgroup
compression and such like it should. And firefox's settings are gone.

(From your later post it sounds as if your memory is definitely bad, or
at the very least not making proper contact.)

So is this the vulnerabilities of fat32? I remember fat16 and it was
terrible. I converted to ntfs. Maybe that will be a better option.

Maybe fat32 does have some weaknesses. Am I experiencing them here?
[]
As to Fat32 vs NTFS


I have experience here and can assure you NTFS has much better fault
tolerance characteristics.

If true, not necessarily a good thing - since things can get into a
worse state before you become aware anything is wrong (and they can then
fall over suddenly, and possibly irretrievably).

I've seen fat systems be rendered useless by writing the data to a slew
of "chk" files

I _think_ you mean by it (them) writing the chk files to contain
"orphaned" data they find. [The _user_ can write chk files as much as
they like.] What NTFS systems do when they find orphaned data, I'm not
sure; maybe they never do find such.

--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Facts do not cease to exist because they are ignored.

Bill Cunningham wrote:
"Bob F" wrote in message
...
First thing to do is test your memory with something like
memtest86+. Bad memory can cause all kinds of problems.

It froze up while doing things. Test 5 had 223 errors and test 7 had 1
error. idk what that means.


It likely means that a memory problem IS your problem. Normally I would run that
test several passes. Any error would tell me to replace the problem memory. You
can just remove the defective memory only 1 (of 2 or more) is the problem, and
your computer should run OK but slower until you replace it. I'd test each
memory in the computer by itself - hopefully just one has a problem.

As far as the static danger that Paul talks about, I just make a point of
touching bare metal of the case of the computer with some part of my body before
and if I can while I am messing around in it. Best not to wear wool or synthetic
clothing while doing it.

"Bob F" wrote in message
...

....

As far as the static danger that Paul talks about, I just make a point of
touching bare metal of the case of the computer with some part of my body
before and if I can while I am messing around in it. Best not to wear wool
or synthetic clothing while doing it.

I thought maybe simply holding onto the chasis would prevent electrical
damage of the ICs. Well I have a 1 Mega Ohm resistor in a wrist strap
somewhere. Are DIMMS all I can use? No SDIMMS or anything newer?

Bill

On Wed, 21 Oct 2015 20:32:53 +0100, "J. P. Gilliver (John)"
wrote:

In message , philo
writes:
On 10/20/2015 12:49 PM, Bill Cunningham wrote:
I have been having some of the same old problems lately with my computer
that I have had off and on for a few years now. Well it's old. I'd say the
ram is going bad. They're DIMMs. but my computer decides it will restart
itself. over and over. Well I lost a file. And OE is not resetting newsgroup
compression and such like it should. And firefox's settings are gone.

(From your later post it sounds as if your memory is definitely bad,




May I suggest that we all say "RAM" instead of "memory"? I had to read
this a couple of times before I understood what you meant. g



or
at the very least not making proper contact.)

So is this the vulnerabilities of fat32? I remember fat16 and it was
terrible. I converted to ntfs. Maybe that will be a better option.

Maybe fat32 does have some weaknesses. Am I experiencing them here?
[]
As to Fat32 vs NTFS


I have experience here and can assure you NTFS has much better fault
tolerance characteristics.

If true, not necessarily a good thing - since things can get into a
worse state before you become aware anything is wrong (and they can then
fall over suddenly, and possibly irretrievably).

I've seen fat systems be rendered useless by writing the data to a slew
of "chk" files

I _think_ you mean by it (them) writing the chk files to contain
"orphaned" data they find. [The _user_ can write chk files as much as
they like.] What NTFS systems do when they find orphaned data, I'm not
sure; maybe they never do find such.

--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Facts do not cease to exist because they are ignored.

Bill Cunningham wrote:
"Bob F" wrote in message
...

...

As far as the static danger that Paul talks about, I just make a point of
touching bare metal of the case of the computer with some part of my body
before and if I can while I am messing around in it. Best not to wear wool
or synthetic clothing while doing it.

I thought maybe simply holding onto the chasis would prevent electrical
damage of the ICs. Well I have a 1 Mega Ohm resistor in a wrist strap
somewhere. Are DIMMS all I can use? No SDIMMS or anything newer?

Bill

You have to look at the user manual and memory section,
to see whether it uses unbuffer (UDIMM) or registered (RDIMM).
You can't mix those two types, so whatever you fill the machine
with must be consistent. For desktops, UDIMM is the most
common type.

Usually Crucial.com has a lookup page, where you can select the machine
make and model number, to find compatible products. Which is another
way to do it. Crucial (Micron) sometimes carries memory families
that Kingston has stopped selling, which is good for moderately old
equipment.

I've had most of my memory problems with "unbranded" memory.
The module itself is made by an unknown maker. The chips
may be brand name, or the chips may be blank. For example,
I bought (8) 512MB generic DIMMs with "MicroQ" (not Micron)
brand memory (never heard of them), and five of the modules
have failed. So while I got the memory for "half price", I
certainly did not get a bargain. I'm quite certain they will
all fail eventually.

Out of all the Crucial I bought, only one stick (Ballistix) failed.
The non-enthusiast RAM was always good (SDRAM for the 440BX). Kingston
here has had a similar good history, with no failures at all.
Kingston is what is running in this computer right now. But I've
had three lots of generic modules that had multiple failures. Typically
on my generic, it fails 1.5 years after purchase.

*******

And the 1 megohm resistor is important, as it slows static discharge
and reduces current density in affected semiconductor junctions. It's
part of safely bringing subassemblies to the same electrostatic
potential. Bare wire and "hard" antistatic connections (no resistor
in the path) are not quite as good. This is why antistatic plastic
is a "fair conductor" not a "good conductor". To slow the rate
of discharge if it happens.

Paul

Paul wrote:
Bill Cunningham wrote:
"Bob F" wrote in message
...

...

As far as the static danger that Paul talks about, I just make a point
of
touching bare metal of the case of the computer with some part of my
body
before and if I can while I am messing around in it. Best not to wear
wool
or synthetic clothing while doing it.

I thought maybe simply holding onto the chasis would prevent
electrical
damage of the ICs. Well I have a 1 Mega Ohm resistor in a wrist strap
somewhere. Are DIMMS all I can use? No SDIMMS or anything newer?

Bill

You have to look at the user manual and memory section,
to see whether it uses unbuffer (UDIMM) or registered (RDIMM).
You can't mix those two types, so whatever you fill the machine
with must be consistent. For desktops, UDIMM is the most
common type.

Usually Crucial.com has a lookup page, where you can select the machine
make and model number, to find compatible products. Which is another
way to do it. Crucial (Micron) sometimes carries memory families
that Kingston has stopped selling, which is good for moderately old
equipment.

I've had most of my memory problems with "unbranded" memory.
The module itself is made by an unknown maker. The chips
may be brand name, or the chips may be blank. For example,
I bought (8) 512MB generic DIMMs with "MicroQ" (not Micron)
brand memory (never heard of them), and five of the modules
have failed. So while I got the memory for "half price", I
certainly did not get a bargain. I'm quite certain they will
all fail eventually.

Out of all the Crucial I bought, only one stick (Ballistix) failed.
The non-enthusiast RAM was always good (SDRAM for the 440BX). The Kingston
here has had a similar good history, with no failures at all.
Kingston is what is running in this computer right now. But I've
had three lots of generic modules that had multiple failures. Typically
on my generic, it fails 1.5 years after purchase.

Wow! So I'm guessing this is due to some surface imperfections? I wonder
how a company that produces "such" can stay in business? Eventually you'd
think buyers (even third party) would catch on, and the word would get out.
So is this "simply" a quality control issue on their part?

I believe even a quality IC manufacturer has a certain failure rate for
their memory ICs, and presumably has some sort of tests that help test it
for some potential failures (like maybe a current leakage test, or something
along those lines, under various temperature conditions), so as to help
predict this. I recall that reliability engineers use the term MTBF
(amongst others) to help quantify some of this, although this is still only
statistically based figure, and not a guarantee for each chip going out the
door.

Bill in Co wrote:
Paul wrote:
Bill Cunningham wrote:
"Bob F" wrote in message
...

...

As far as the static danger that Paul talks about, I just make a point
of
touching bare metal of the case of the computer with some part of my
body
before and if I can while I am messing around in it. Best not to wear
wool
or synthetic clothing while doing it.
I thought maybe simply holding onto the chasis would prevent
electrical
damage of the ICs. Well I have a 1 Mega Ohm resistor in a wrist strap
somewhere. Are DIMMS all I can use? No SDIMMS or anything newer?

Bill
You have to look at the user manual and memory section,
to see whether it uses unbuffer (UDIMM) or registered (RDIMM).
You can't mix those two types, so whatever you fill the machine
with must be consistent. For desktops, UDIMM is the most
common type.

Usually Crucial.com has a lookup page, where you can select the machine
make and model number, to find compatible products. Which is another
way to do it. Crucial (Micron) sometimes carries memory families
that Kingston has stopped selling, which is good for moderately old
equipment.

I've had most of my memory problems with "unbranded" memory.
The module itself is made by an unknown maker. The chips
may be brand name, or the chips may be blank. For example,
I bought (8) 512MB generic DIMMs with "MicroQ" (not Micron)
brand memory (never heard of them), and five of the modules
have failed. So while I got the memory for "half price", I
certainly did not get a bargain. I'm quite certain they will
all fail eventually.

Out of all the Crucial I bought, only one stick (Ballistix) failed.
The non-enthusiast RAM was always good (SDRAM for the 440BX). The Kingston
here has had a similar good history, with no failures at all.
Kingston is what is running in this computer right now. But I've
had three lots of generic modules that had multiple failures. Typically
on my generic, it fails 1.5 years after purchase.

Wow! So I'm guessing this is due to some surface imperfections? I wonder
how a company that produces "such" can stay in business? Eventually you'd
think buyers (even third party) would catch on, and the word would get out.
So is this "simply" a quality control issue on their part?

I believe even a quality IC manufacturer has a certain failure rate for
their memory ICs, and presumably has some sort of tests that help test it
for some potential failures (like maybe a current leakage test, or something
along those lines, under various temperature conditions), so as to help
predict this. I recall that reliability engineers use the term MTBF
(amongst others) to help quantify some of this, although this is still only
statistically based figure, and not a guarantee for each chip going out the
door.

I don't know if it's possible to sort ICs by IDDD any more.
That was the technique at one time, where it was known that
simple CMOS logic gates had "zero" current flow when not toggling.
If you found a chip with a measurable current flow, that was
a sign it was defective. You could do that sort of analysis at
wafer-sort, and mark as defective, some of the dice, before
sawing up the wafer and going to the trouble of packaging
the chips. Modern chips have a measurable leakage, which makes
IDDD a dubious test to run. It was certainly a nice test,
while it lasted.

I expect a lot of these failures are "chemistry". At one time,
there was an IBM web page, that detailed a contamination issue
on one of their fab production lines. It turned out that
chlorine contamination was detected, and they described all
the steps it took them to find it. We had a problem at our
own fab once, where the fab was "zero yield" for at least a month,
and that was a chemistry issue too. Some of those can last
for months, if they can't find the issue in a cursory examination.
It can take up to 12 weeks to manufacturer a chip, and maybe
70 to 130 separate manufacturing steps (sputtering, ion
implantation, doping, whatever). So there are lots of opportunities
for something dirty to get in there and screw things up.

This is why, if you look at say, an Intel fab, materials are
carried from station to station by robots. We used to use humans
on our line. They're high maintenance and need a lot of
lubricating with coffee. Our fab was not a serious effort, it did
produce relatively small quantities of parts, and at least half
the effort was research. They did eventually make some nice stuff,
but the company management got a clue and cut the whole
operation. The idea being, you just go to TSMC and get them to
make your chips. AMD has kinda learned the same thing, only
in their case their hand was forced by economics.

Paul

"Paul" wrote in message
...

....

I'm just a little confused as to why a file. Like the one on my HD
disappeared. It wasn't open. Why would it have been in memory? It simply
vanished.

Bill