Undeletable file. I'm stumped.

[BillW50]

In glee typed:

snip
I already read and responded to the comment you related being made by
Paul.... he was referring in that comment to modified specialized Live
CDs like Kaspersky's rescue CD, which as I already stated is NOT a
standard Linux Live CD.

Paul did talk about that. But he also talked about this from the get-go:

[Thursday, March 24, 2011 8:48 PM]
The only practice I don't approve on, from the Linux community,
is "scanning" of drives as part of the startup sequence. Some
LiveCD distros, are known to "search" for a copy of the image
you're booting from. Presumably the purpose, is to do a
loopback mount of the image, as a replacement for accessing the
CD itself. But I still don't approve of monkey-business. A
LiveCD should just mind its own business. ~ Paul

I note that while you give a quote you say is from Paul, I have not
found that post online... do you have a link to his actual full post
in an archive such as Google Groups?

Under Google News, it led me here. It's the second from the end.

http://www.pcreview.co.uk/forums/windows-not-load-t4036245p2.html
snip

Other than the one assertion apparently made by Paul, I have found no
documented evidence that any *standard* Linux Live CD tries to run from
the ISO image if found on the hard drive. Ping Paul and ask him to
provide evidence of this.

Yes, that link shows the quote you've been using. Again, there is no
evidence given. Paul is a very smart guy, but his claim that by using
the Linux 'top' command he is somehow showing that the Windows page file
is being used by the Live CD is an incorrect assumption on his part. The
'top' command shows CPU processes, kind of like a task manager. Paul
stated he used 'top' to see how much swap was evident (it does NOT show
where the swap file is), and compared it in size to the existing page
file(s) on the Windows and other partitions.... and found similar sizes.
That is in no way evidence that the Linux Live CD is using the Windows
page file. It shows that the Linux Live CD, running on a system with X
amount of RAM, will allocate a certain swap size, similar in size to
what Windows allocates on the same system.

The fact is, a Live CD allocates a virtual swap file in RAM.... that is
what he is seeing with the 'top' command. The Live CD boot divides the
RAM into segments, and creates a virtual swap file in RAM from one
segment. The another segment created from RAM is used as "storage" like
a virtual hard drive, the remaining RAM is used like standard RAM for
loading programs, etc. Those Live CDs (Gentoo included) use squashfs to
accomplish this.

I've been looking through the archives all day that led me to the
Windows swapfile. Paul wasn't the first who mentioned it. I did find
some posts of this problem from 2009 which I said then that restoring
the registry with ERUNT also corrected the problem. I didn't recall that
part until I read it today.

Anyway the post that I am looking for was by some Linux guru from 2009 I
believe that knew tons about Linux and he stated he thought Ubuntu
creating a swapfile on the Windows drive was the problem. I just haven't
found his post so far. But I'll keep looking.

 

[BillW50]

...and just to be clear, as a P.S. to my other reply, if you have an
actual Linux installation on the hard drive, in addition to Windows or
instead of Windows, some distros of LIve CD *may* use the existing swap
partiton of the installed Linux system. This is not the same as using
the Windows page file, which the Live CDs do *not* do.

In this discussion, we're not talking about a computer that has a Linux
installation on the hard dirve, though.... we are talking about systems
with Windows only. The hard drive is not mounted without specific input
from the user, when using the defaults of the Live CD.

I wouldn't bother mixing Windows and Linux on the same machine personally.

 

[BillW50]

IBM rightly ignored your vague declaration that "omg something is
wrong" and again rightly did not give you any credit because you didn't
actually contribute anything that would help identify the problem or
provide a solution.

Among other things, I've done software quality assurance testing and
what you are doing doesn't qualify. Documentation, debug logs, screen
captures, before and after disk images / file CRCs, etc., verification
on multiple systems, systematic elimination of other variables to
determine the actual cause of a problem that was observed - those are
(some of) the things that prove a problem. Until and unless you
provide a repeatable set of steps to duplicate the issue, you are just
waving your hands and proclaiming that the sky is falling. Pick up the
acorn, Henny Penny, and recognize it for what it is. In this case, it
is a consequence of your own actions, not a flaw in the system.

How long have you been doing this? You can't send in a bug report just
saying "omg something is wrong". That just doesn't fly. There are tons
of other information that you have to supply. Hardware, build, drivers,
open applications, logs, etc. I sent all of that in over a dozen times
over that period of time.

 

[BillW50]

IMHO, Bill's pretty much pi**ing into the wind with his claims until
he steps up and provides some actual technical information. A blinking
HD activity LED doesn't tell a convincing story.

How about something like, a complete set of file CRCs from before and
after running the live CD. Compare the CRCs to see which, if any,
files have changed. A second step could be a DIFF to see how the
affected file(s), if any, have changed.

I suggest doing this by booting a live CD twice, with CRC snapshots
taken before, after, and after, since booting Windows will absolutely
cause changes that will skew the test results. We could discuss the
finer points until everyone's satisfied that it's a valid test, but in
the meantime we've basically got the resident nut job making a claim
and not being able to back it up. It's a stalemate until someone steps
up and does some *actual* testing, rather than watching a flickering
LED.

Oh Ubuntu Live is accessing the Windows drive way too much to be
nothing. I'm not surprised that it doesn't mean anything to you. ;-)

 

[Char Jackson]

Oh Ubuntu Live is accessing the Windows drive way too much to be
nothing. I'm not surprised that it doesn't mean anything to you. ;-)

It shouldn't mean anything to you, either.

Get some *real* data, please. See Zaphod's post for some excellent
tips.

 

[Char Jackson]

How long have you been doing this? You can't send in a bug report just
saying "omg something is wrong". That just doesn't fly. There are tons
of other information that you have to supply. Hardware, build, drivers,
open applications, logs, etc. I sent all of that in over a dozen times
over that period of time.

If you're so experienced, why are you only giving us a flickering HD
activity LED?? Until you produce some actual test results, you've got
nothing.

 

[Char Jackson]

I've been looking through the archives all day that led me to the
Windows swapfile.

So that's where you've been all day.

Paul wasn't the first who mentioned it. I did find
some posts of this problem from 2009 which I said then that restoring
the registry with ERUNT also corrected the problem. I didn't recall that
part until I read it today.

Are you now going to say that your live CD wrote to the Windows
Registry? This just keeps getting better.

Anyway the post that I am looking for was by some Linux guru from 2009 I
believe that knew tons about Linux and he stated he thought Ubuntu
creating a swapfile on the Windows drive was the problem. I just haven't
found his post so far. But I'll keep looking.

I think your time would be better spent generating some actual test
data. Dredging up an old post isn't going to win you any points.

 

[Zaphod Beeblebrox]

If you're so experienced, why are you only giving us a flickering HD
activity LED?? Until you produce some actual test results, you've got
nothing.

+1

 

[Zaphod Beeblebrox]

How long have you been doing this?

It isn't how long you have been doing something, it is how well you do
it. Clearly, you fail that measurement of success.

You can't send in a bug report just
saying "omg something is wrong". That just doesn't fly. There are tons
of other information that you have to supply. Hardware, build, drivers,
open applications, logs, etc.

None of which matter a whit if the description of the problem is
lacking and you don't supply a repeatable set of steps to duplicate the
problem. Submitting a report that "omg something is wrong, here are my
specs" doesn't cut it.

I sent all of that in over a dozen times
over that period of time.

Then apparently what you sent was either inaccurate or insufficient or
both. Likewise, if you presented the information in the same
disorganized and anecdotal way you have here and with the same
attitude, it is likely what you submitted was dismissed out of hand.
See above: you are doing it wrong. NB: There are classes that can
teach you how, assuming you are willing and capable of learning.

--
Zaphod

Pan-Galactic Gargle Blaster: A cocktail based on Janx Spirit.
The effect of one is like having your brain smashed out
by a slice of lemon wrapped round a large gold brick.

 

[BillW50]

In Zaphod Beeblebrox typed:

+1

Sure no problem. I charge $150 per hour. So where do I send the bill
too? For free, I tell you that there is something wrong here and it
needs to be investigated.

 

[BillW50]

In Zaphod Beeblebrox typed:

It isn't how long you have been doing something, it is how well you do
it. Clearly, you fail that measurement of success.


None of which matter a whit if the description of the problem is
lacking and you don't supply a repeatable set of steps to duplicate
the problem. Submitting a report that "omg something is wrong, here
are my specs" doesn't cut it.


Then apparently what you sent was either inaccurate or insufficient or
both. Likewise, if you presented the information in the same
disorganized and anecdotal way you have here and with the same
attitude, it is likely what you submitted was dismissed out of hand.
See above: you are doing it wrong. NB: There are classes that can
teach you how, assuming you are willing and capable of learning.

You don't understand the scope of the problem. I used copy and paste to
and from OS/2 and DOS all of the time back then. This left the OS in an
unstable state. And it would take about a couple of more days before the
OS would actually lockup. By then there was seemingly no rhyme or reason
for the OS lockup. After all, it had taken IBM two years just to find
it. I am sure I was not the only one who found this bug, as OS/2 had
millions of users back then. Although I was the only one who spoke
publicly about it. And I have enough experience that I knew it wasn't a
problem with my hardware or me.

 

[Zaphod Beeblebrox]

In Zaphod Beeblebrox typed:

You don't understand the scope of the problem. I used copy and paste to
and from OS/2 and DOS all of the time back then. This left the OS in an
unstable state. And it would take about a couple of more days before the
OS would actually lockup. By then there was seemingly no rhyme or reason
for the OS lockup. After all, it had taken IBM two years just to find
it. I am sure I was not the only one who found this bug, as OS/2 had
millions of users back then. Although I was the only one who spoke
publicly about it. And I have enough experience that I knew it wasn't a
problem with my hardware or me.

Ah yes, the old "I know what I'm doing, it can't be me, it must be the
software" method of SQA. How quaint. I'm sure you used the equally
impressive "I sent you that bug report with stream of consciousness
anecdotal evidence of a problem at least a hundred times, what do you
mean nobody else is reporting it and you can't reproduce it" approach
to documenting and reporting the issue. <slow clap> Bravo! I must say,
you are indeed in a class all your own.

--
Zaphod

Adventurer, ex-hippie, good-timer (crook? quite possibly),
manic self-publicist, terrible bad at personal relationships,
often thought to be completely out to lunch.

 

[Achilles]

On 09/12/2012 07:41 AM, Zaphod Beeblebrox wrote:

Live Cds for Ubuntu and Mint will automatically use a swap partition if
they see a linux swap partition on the local disk. No harm could ever
come from this.

 

[glee]

On 09/12/2012 07:41 AM, Zaphod Beeblebrox wrote:

Live Cds for Ubuntu and Mint will automatically use a swap partition
if they see a linux swap partition on the local disk. No harm could
ever come from this.

Yes, already mentioned.... but that is not the Windows page file, which
is what BillW50 *claims* a Linux Live CD uses.

 

[BillW50]

In glee typed:

Yes, already mentioned.... but that is not the Windows page file,
which is what BillW50 *claims* a Linux Live CD uses.

Actually I am just the messenger. I am still looking for that post that
someone told me what Linux can do to a Windows install (including using
the Windows swapfile). I thought it was back in 2009 on a newsgroup
after I told my horror story with Ubuntu Live.

I did find another post on the Dell newsgroup back in August of 2011
which I totally forgotten about. As Monica's Windows started failing to
boot. Somebody mentioned to use Ubuntu Live to pull off her important
files off of the drive. I warned Monica not to use Ubuntu, as Ubuntu
Live burned me with a Windows drive before. She didn't listen and Ubuntu
Live read her hard drive and then later toasted it where nothing could
read the drive after that.

Still looking for that post. maybe he told me on the eeeuser forum or
something and I am looking in the wrong place, i.e. newsgroups. :-(

 

[Char Jackson]

In glee typed:

Actually I am just the messenger. I am still looking for that post that
someone told me what Linux can do to a Windows install (including using
the Windows swapfile). I thought it was back in 2009 on a newsgroup
after I told my horror story with Ubuntu Live.

I did find another post on the Dell newsgroup back in August of 2011
which I totally forgotten about. As Monica's Windows started failing to
boot. Somebody mentioned to use Ubuntu Live to pull off her important
files off of the drive. I warned Monica not to use Ubuntu, as Ubuntu
Live burned me with a Windows drive before. She didn't listen and Ubuntu
Live read her hard drive and then later toasted it where nothing could
read the drive after that.

Still looking for that post. maybe he told me on the eeeuser forum or
something and I am looking in the wrong place, i.e. newsgroups. :-(

What usually happens in cases like this is, if you keep up your
search, you'll eventually find what you've been looking for and you'll
present it to the group in a grand 'A Ha!' and 'Ta Da!' moment. Next,
someone will point out that it doesn't actually say what you thought
it said, and therefore it hurts your case rather than helps it. And
around and around we go, with the cycle starting again at the
beginning.

 

[Zaphod Beeblebrox]

What usually happens in cases like this is, if you keep up your
search, you'll eventually find what you've been looking for and you'll
present it to the group in a grand 'A Ha!' and 'Ta Da!' moment. Next,
someone will point out that it doesn't actually say what you thought
it said, and therefore it hurts your case rather than helps it. And
around and around we go, with the cycle starting again at the
beginning.

+1

 

[BillW50]

In Zaphod Beeblebrox typed:

+1

Apparently you two don't know me too well. As I am more interested in
the truth and I could careless if I was wrong about something. Nor am I
afraid to admit that I do make mistakes from time to time.

Now Philo says that Linux can't do anything to the Windows drive unless
you mount it first. I told Philo I did not mount the Windows drive. Now
Joe Internet Jun 24 '11 at 17:38 says:

Most Linux live-cds boot into a desktop environment, so you
just open the file manager and delete your files. Not all,
however, automatically mount NTFS partitions. Two that I know
of that do mount them are Ubuntu and Centos.

http://superuser.com/questions/3016...d-hiberfile-sys-on-secondary-disk-old-windows

If true, that solves that problem. As I didn't mount the Windows drive,
but Ubuntu did.

Now Char mentioned me stating that I found an old post of mine that said
restoring the Windows registry with ERUNT also fixed Windows XP after
Ubuntu Live corrupted it somehow. Yes that is puzzling to me too. As one
wouldn't think that Ubuntu would be messing around the Windows registry
for any reason.

I'm still looking for the post where that guy know told me what he
thinks what went wrong was Ubuntu created a Windows swapfile to use for
itself. Until I find it, I have to rely on my memory.

So I have a theory that might put everything together. If Ubuntu
automatically mounts NTFS like Joe Internet says. And my understanding
is that Ubuntu will grab a Linux swap partition if it can find one. If
it can't, then according to the post I am looking for says Ubuntu will
seek out a Windows pagefile and use that one. And according to that
post, if Windows doesn't have one, Ubuntu will create one there.

Now if the above is true, I think I know what happened. Windows now
tries to boot, finds a swapfile and the registry says no swapfile.
Windows gets confused and makes changes to Windows registry and deletes
the swapfile. Which results into an unbootable Windows.

TO BE CONTINUED.

 

[Zaphod Beeblebrox]

Apparently you two don't know me too well.

On the contrary, we know you all too well.

As I am more interested in the truth and I could careless
if I was wrong about something. Nor am I afraid to admit
that I do make mistakes from time to time.

And then you follow with the most convoluted and least likely set of
circumstances that I think I've ever seen. And I might add, one that
could be reduced to a simple scenario that could be tested in about 2
minutes if you had half as much expertise as you claim. (I'll leave it
as an exercise for the class to determine what that test would consist
of.) Having already performed that test on several spare systems I
have at my disposal, I can tell you that you are wrong. Not that this
is a surprise, but there it is. I await your admission that you are
wrong about this, but I won't hold my breath - you've never admitted
you were wrong about anything of consequence in these newsgroups, and
given your personality I sincerely doubt that will change now.

--
Zaphod

Pan-Galactic Gargle Blaster: A cocktail based on Janx Spirit.
The effect of one is like having your brain smashed out
by a slice of lemon wrapped round a large gold brick.

 

[BillW50]

On the contrary, we know you all too well.


And then you follow with the most convoluted and least likely set of
circumstances that I think I've ever seen. And I might add, one that
could be reduced to a simple scenario that could be tested in about 2
minutes if you had half as much expertise as you claim. (I'll leave it
as an exercise for the class to determine what that test would consist
of.) Having already performed that test on several spare systems I
have at my disposal, I can tell you that you are wrong. Not that this
is a surprise, but there it is. I await your admission that you are
wrong about this, but I won't hold my breath - you've never admitted
you were wrong about anything of consequence in these newsgroups, and
given your personality I sincerely doubt that will change now.

I am not wrong about running Ubuntu Live and it ruins Windows and makes
it unbootable. That I am positive about. The exact conditions that
causes this, I am not sure about. I also never seen it happen from
Ubuntu Live CD yet (although Monica has). I personally have just seen it
happen on a flash.

Testing it in 2 minutes? Naw... I even have the very same flash card
that I used for Ubuntu Live. Although it has BartPE on it now. Plus I
need to back it up and then use unetbootin-windows-299.exe to image
ubuntu-eee-8.04.1.iso on the exact same flash. Then run Ubuntu Live from
the flash and it will toast the Windows XP install on the EeePC.