Any way to TOTALLY disable Windows 98 Disk Cleanup tool?

Lostgallifreyan wrote in
:

I usually see reports of space being
wrong after bad shutdown, fixed by Scandisk subsequently, but the only
file usually harmed by this is the one just written. As it's the one I
was most recently working with, I usually have immediate ways to restore
it because I always save twice or more to separate locations..


Actually I have often checked a file after Scandisk has done its bit, the
file is nearly always ok, it's just the free-space count that didn't get
updated during the shutdown. Next worst case is also usually THE worst case,
some lost clusters, likely during a crash so severe that I forced power off,
so a disk controller didn't get to flush its cache. I usually do this when a
browser got hijacked by excess data and forced the OS into seizure. At that
point, the last thing I want is that data, I just force control back on the
machine because as I see it, when some remote site has more control than I
do, even for an instant, that's a moment of great risk. I'd rather force it
that sit there wondering what's going on at my expense..

On 7/8/2012 14:04, Lostgallifreyan wrote:
Bill wrote in
ng.com:

Right... at the higher level a dir is the container for the file and its
data, and can be considered equal to the size of its contents, but at
the low level a dir is just a container for file entries. The data
clusters are a separate construct. The main point I was trying to make
is that you don't need to reserve space other than what is needed for
the directory itself. The data clusters are already allocated for every
file that will be deleted.


They are, but until truly deleted, they are part of the reckoned size (so as
not to encroach on even more...). I think of the recycle bins as 'move'
operations done on same disk.

Exactly. It's a move operation. And is why there's a recycled directory
on each volume. You can't move across drives. That would be a copy and
then del op.

About higher level dirs, I never explored very deeply, but I have seen that
chunks of file tables appear at other locations. I think the structure is the
same, being clusters containing pointers to files. As a directory is a file,
differing only in its purpose of holding data that points to other files (and
subdirs) I imagine it can effectively point to other parts of itself if it
were scattered across discontigous clusters, as with any other file.

Right, a dir can link to discontigous clusters.

But as I
say, I never looked that deep so I'm not sure. What I am sure of, is that the
need for a bit of space to link what might otherwise become broken FAT
chains, is maybe why there is a monitor watching for stuff than might
conflict with that need, or any other need for that bit of space.

The only need for the dir to grow is if it has filled to capacity with
file entry data, in which case the dir/file would then add and link to
another cluster. The FAT is different. It's all allocated during the
format with room for a pointer to every cluster.


That said,
I don't see much trouble on my system, I usually see reports of space being
wrong after bad shutdown, fixed by Scandisk subsequently, but the only file
usually harmed by this is the one just written. As it's the one I was most
recently working with, I usually have immediate ways to restore it because I
always save twice or more to separate locations..

The free space count being out of whack is trivial, and happens quite a
bit. Though in 98Guy's case, (if that is the problem), it's far out of
whack. I'd definitely do a back up before running a scandisk on a drive
with that many clusters.

On 7/8/2012 14:06, Lostgallifreyan wrote:
"J. P. Gilliver wrote in
:

It's not wrong without context. See you in APIHNA ... (-:


If I let myself take that excuse, the ravening hoardes will be fast led
across the Earth by the four horsemen of the Apocalypse, and all will be SO
not right with the world that I dare not.... What is APIHNA? API Hell than
kNows no Abounds? If so, been there, and burned the t-shirt.

A Place In Hell with No Airconditioning?

On Sun, 08 Jul 2012 20:12:55 +0100, Bill Blanton
wrote:

On 7/8/2012 14:06, Lostgallifreyan wrote:
"J. P. Gilliver wrote in
:

It's not wrong without context. See you in APIHNA ... (-:


If I let myself take that excuse, the ravening hoardes will be fast led
across the Earth by the four horsemen of the Apocalypse, and all will
be SO
not right with the world that I dare not.... What is APIHNA? API
Hell than
kNows no Abounds? If so, been there, and burned the t-shirt.

A Place In Hell with No Airconditioning?

All 'postrophe's are happy nearly anywhere.

--
[dash dash space newline 4line sig]

Albi CNU

Bill Blanton wrote in
ng.com:

The free space count being out of whack is trivial, and happens quite a
bit. Though in 98Guy's case, (if that is the problem), it's far out of
whack. I'd definitely do a back up before running a scandisk on a drive
with that many clusters.


Indeed. That kind of backup is the fastest way to get a defragged copy too,
which can help if data recovery is ever needed. I do this kind of copy every
time I re-organise large stores of large files. (I also batch-script it so
they are laid in proper order, as Explorer tends to copy (for example,
eight files) 12345678 resulting in 82345671 or 81234567 for reasons best
known to itself. I also create all the dirs first, so that every file pointer
exists at the start of the archive disk. It took ahile to figure out all that
and create the scripts, but all subsequent use makes it almost as easy as
pushing a single button. As well as good backupos, I never need defrag,
because once I copy one way, I check MD5's and copy back onto a wiped
partition, so both are done overnight by the time I wake up. I only do this
about twice a year, but it's enough.)

Bill Blanton wrote in
ng.com:

The only need for the dir to grow is if it has filled to capacity with
file entry data, in which case the dir/file would then add and link to
another cluster. The FAT is different. It's all allocated during the
format with room for a pointer to every cluster.


Ok. Even so, that would pay to have some watch on disk space. But I never had
serious trouble with it reaching the last byte (excluding unusable slack
space) before shouting about it, so I think wanting that last 3% is fine. I
can't remember exactly when and why I stopped seeing that report, but I think
98-Lite had more to do with it than anything else.

One other possibility is that the W95 Explorer.exe and Shell32.dll I use
don't have that free space warning. I know that long past, I used to see
prompts for low space to be cleaned up by the cleanup manager, and that it
can definitely be removed either by the shell swap, or removing the cleanup
manager, and setting recycle bins to zero. I looked again at TweakUI, but I
don't think that has control over this.

One small oddity is that if a recycle bin was itself deleted, and a drive
gets full to the last byte, and I delete something to make space to retry,
the recycle bin fails to gets its icon because at the moment of delete,
before I confirm, that bin IS made, but its internal Desktop.ini is not
because there was no space for it.

Bill Blanton wrote in
ng.com:

A Place In Hell with No Airconditioning?



Excellent. Such is the UK climate right now that I think I'm already there.

"Harry Vaderchi" wrote in newsp.wg46j1ov1r0rdn@dell3100:

All 'postrophe's are happy nearly anywhere.


Careful now... That way lies unravelling of the space-time continuum.

You still looking for a solution?
Try copy and paste, make a REG file:

REGEDIT4

[HKEY_CURRENT_USER\Software\Microsoft\Windows\Curre ntVersion\Policies\Explor
er]
"NoLowDiskSpaceChecks"=dword:00000001

[HKEY_USERS\.DEFAULT\Software\Microsoft\Windows\Cur rentVersion\Policies\Expl
orer]
"NoLowDiskSpaceChecks"=dword:00000001

[HKEY_LOCAL_MACHINE\System\CurrentControlSet\Contro l\FileSystem]
"DisableLowDiskSpaceBroadcast"=dword:00000210

Sorry, crappy Outlook Express messed up my previous entries .. you know how
to fix them.