Prev: DPC Latency Checker
Next: alt.comp.freeware links at Sat Jul 10 21:20:01 2010
From: Bob Adkins on 12 Jul 2010 14:28
On Mon, 12 Jul 2010 10:26:46 -0500, VanguardLH <V(a)nguard.LH> wrote:


>And how does that actually do anything to help the user of Windows? It
>doesn't.


It does!

I can't count the times I have removed error message boxes that pop
up. Yes,,, you can do it by hand. Sometimes you find it in 5 minutes,
but if you are working on an old, slow computer, it can take hours. I
rather unleash a registry cleaner on it.


> Orphaned entries do not slow access to the registry because

They do if they try to load a DLL or EXE that's no longer there!

>the registry is copied into memory - and memory is random-read so the
>orphaned entries do not slow accessing any parts of the registry.

That's fine if you have 2gb or more memory, but half my customers are
trying to run XP on 512 or less, or Vista on 1gb or less! Not only
that, but old slow HDD's that the memory has to cache out to.

>Actually I've seen many registry cleaner that don't even tell the user
>what actions they propose to make - and the ignorant users you mentioned
>above won't know if the cleaner should make those changes.

Just how many ignorant users do you think there are on ACF? OK, I'm
one, but can you name me another one? :)

Look, we're seasoned users here. People that use UseNet are special,
and you don't have to coddle them.


>I'm not arguing that registry cleaners should never be used. They make
>a nice automated tool to do all the work for you but YOU need to be the
>final authority as to what should actually get changed.

Where do you get off with those outrageous statements? I violently
agree! :)
From: Bob Adkins on 12 Jul 2010 14:35
On Mon, 12 Jul 2010 10:32:09 -0500, VanguardLH <V(a)nguard.LH> wrote:


>I believe Softpedia's policy is that these options are allowed in
>non-adware BUT it must be disabled by *default*. If Softpedia is
>listing this as adware is because the installer has these options
>ENABLED by default.


I think that's a great policy. That's why I DL from SoftPedia. They
ride herd on Adware and Crapware, and they also stay up to date. Their
screen shots are usually better than the developer's site, and so is
their download speed.

> It's the same reason why Softpedia classified
>Comodo's Internet Suite as adware, Comodo complained, Softpedia stuck to
>their guns regarding their adware definition, so Comodo requested that
>Softpedia not list Comodo's software anymore and Softpedia complied
>(since they won't force anyone to have their product listed on their
>site).

Wasn't that a thing of beauty? Comodo had a very nice security suite,
but was throwing the baby out with the bath water. SoftPedia stuck to
their guns. Comodo blinked first, and everyone's happy. :)


>There are a lot of installers now that default to "on" for installing
>their foistware. Users need to get in the habit of selecting a custom
>install instead of using the default install as well as reading the

Agreed. I always use custom install and watch for innocent-looking
little check boxes. :)
From: Nicetameetya on 12 Jul 2010 19:15

[Default] On Mon, 12 Jul 2010 08:33:14 -0500, Bob Adkins <me(a)pit.com>
told us in complete confidence:

>On Sun, 11 Jul 2010 18:47:19 -0500, "HeyBub" <heybub(a)NOSPAMgmail.com>
>wrote:
>
>>1. There is absolutely NOTHING good that can come from a registry clearner -
>>not efficiency, not speed, not elimination of crashes or confusion.
>
>That's not quite true. You can remove dozens to hundreds of scraps
>that "uninstalled" programs leave behind. I've helped hundreds of
>people that get error dialogs "can't find so-and-so" when starting
>windows. Rather than spending up to an hour to find out where the
>orphan registry entries are, I run a registry cleaner that takes 5
>minutes. Most have a "conservative mode" that will not tread heavily
>on the registry.
>
>>2. Registry cleaners can render your computer inoperable.
>
>Been experimenting with good to frankly horrible registry cleaners for
>15 years. My computer has never been rendered inoperable, or Windows
>failed to start.

I haven't commented before now because the registry cleaner I've been
using over the past ten years is not freeware. It's called Windows
Registry Repair Pro (from 3B Software) and cost me around twenty
dollars for a license that has entitled me to upgrades since.

I've used WRRP on three different machines over the years, but
currently running it under XPpro on a dual-core Intel 3.4GHz machine
with 2gigs of RAM.

Never had a problem and notice a marked improvement in speed after I
run WRRP and clean my registry every couple of months.

From: VanguardLH on 13 Jul 2010 01:17
Bob Adkins wrote:

> VanguardLH wrote:
>
>>And how does that actually do anything to help the user of Windows? It
>>doesn't.
>
> It does!
>
> I can't count the times I have removed error message boxes that pop
> up. Yes,,, you can do it by hand. Sometimes you find it in 5 minutes,
> but if you are working on an old, slow computer, it can take hours. I
> rather unleash a registry cleaner on it.

That doesn't slow the load of Windows or your login. The prompt doesn't
halt Windows loading or your userprofile loading. What you describe is
a nuisance experienced solely by the user. Nothing of the host is
slowed or suspended.

What is interesting is the opposite effect: the user wondering if
anything is happening because they don't see any progress. An example
is a real-time scheduling system for the airlines that seemed to be
running very slowly. The programmer's immediate fix was to adding
progress messages (i.e., "got here" prompts that programmers often add
just to see something happen and how far they got). They satisfied the
customers that something was happening although nothing had actually
been fixed yet regarding the actual slowness of the software. It shut
up the users to let the programmer work on the real problem in peace.
In your case, you are talking about prompts that nuisance the *user* and
have no effect on Windows itself. Slowing the user by making them waste
time to solve an unknown or unexpected prompt is NOT the same as slowing
the OS.

>> Orphaned entries do not slow access to the registry because
>
> They do if they try to load a DLL or EXE that's no longer there!

That doesn't slow access to the registry. That just nuisances the user
and will slow their *login* time (assuming they take time to act on the
prompt versus Windows which will continue and complete doing its
userprofile load despite the prompt to the user). The startup of
Windows is not slowed by these prompts for no longer existing files.
The login and load of the userprofile is not slowed by these prompts.
Both Windows and the userprofile continuing loading despite the display
of these prompts. There is no dependence in Windows that halts its
progress to load because one startup item is incorrectly defined (which
is NOT just for orphaned registry entries but can also be caused in
incorrent installs, too). What you describe is a nuisance to the user
that takes THEIR time to investigate if they so choose to eliminate the
prompts.

>>the registry is copied into memory - and memory is random-read so the
>>orphaned entries do not slow accessing any parts of the registry.
>
> That's fine if you have 2gb or more memory, but half my customers are
> trying to run XP on 512 or less, or Vista on 1gb or less! Not only
> that, but old slow HDD's that the memory has to cache out to.

Since it is a one-time read during Windows startup to load the registry
into memory, the cache in the HDD is irrelevant. It won't be used again
and it wasn't used in the first place.

How big do you think is the registry? The text you see in the editor is
NOT the content of the registry's database files. The size of exporting
a *text* version of the registry is NOT the size of the registry. That
ends up bloating the size of the output. Go look at the registry .dat
file under your userprofile and the registry database files under the
%windir%\system32\config folder (I'd have to go check again which of
those files are involved in the registry versus the account permissions
saved in the SAM database). The registry files and the memory copy is a
*binary* database tree, not the text you see in an editor.

Most users end up saving all of 5KB to 30KB (and very infrequently
somewhere up to 1MB) when they use a registry cleaner - and that's after
years of using their Windows and installing/uninstalling many programs
during that interim - and that's the text saving and not the binary size
change of the registry files. They reduce the registry's size by very
little but at great risk. The return on investment isn't worth the
risk.

>>Actually I've seen many registry cleaner that don't even tell the user
>>what actions they propose to make - and the ignorant users you mentioned
>>above won't know if the cleaner should make those changes.
>
> Just how many ignorant users do you think there are on ACF? OK, I'm
> one, but can you name me another one? :)
>
> Look, we're seasoned users here. People that use UseNet are special,
> and you don't have to coddle them.

Sorry, a seasoned user in MS Word, SQL, HTML, C++, Perl, Windows admin,
software QA, or other technical pursuits does not equivalence that
expertise with understanding the registry. For example, I have yet to
see one registry cleaner that looks under different hardware profiles to
verify that what it thinks are orphaned entries are actually used when a
different OS environment is loaded (e.g., some diagnostic tools actually
load their own OS but add registry entries but they don't mean anything
until that *other* diag OS gets loaded). These cleaners only know (a
stagnant set of rules) for the current instance of the OS that is
loaded.

>>I'm not arguing that registry cleaners should never be used. They make
>>a nice automated tool to do all the work for you but YOU need to be the
>>final authority as to what should actually get changed.
>
> Where do you get off with those outrageous statements? I violently
> agree! :)

I guess I wanted to make sure that you understood that registry cleaners
can be helpful but they can also be very hazardous even in the hands of
so-called experts. I target shoot at the range for high accuracy for
fun but the same loaded gun in the hands of a child (whether a permanent
mental condition regarding their Windows expertise or a lapse of
judgement by an "expert") is obviously very dangerous. Yet many of the
other shooters that I know of still don't even know how to dismantle
their firearm to do a thorough cleaning or to repair it. When you dig
into the registry, you are at this same level of expertise or harmful
experimentation. If it's a firearm or OS that you don't care about
(i.e., it is for testing and you have a means of immediate recovery)
then go dig in however far you want despite not yet having the expertise
to put it all back together properly. Most users that ask about
registry cleaners are using them on their sole "production" host rather
than on a "test" host.

The biggest problem that I've seen with registry cleaners is not that
there is an immediate problem in loading Windows, in loading your
userprofile, or in the usability or behavior of Windows or your
applications but that the effect isn't exhibited until quite awhile
later so these users don't connect their registry cleanup with the
problem they discover later. They're playing Jenga with their registry:
might not fall immediately but they've been weakening it over time and
then, poof, something very bad happened but it couldn't possibly be the
registry cleaner because they used that a week or a month ago. The bad
effect of the registry might not cause collapse until sometime later.
And collapse doesn't mean Windows won't load anymore. It can be
something simpler, like losing some feature in Windows or within an
application because the cleaner didn't understand the interdependence of
the registry entries. After all, registry cleaners only have a stagnant
set of rules in their code. If they were perfect, there would only be
version 0.1 and never a later version to fix mistakes or update their
rules.

It looks like we are close to being on the same page. I tend to
dissuade users from using registry cleaners because most that ask about
them don't have the expertise to understand if all the proposed changes
should be allowed. You tend to suggest their use but give some warning.
Two viewpoints pointing at the same thing.
From: HeyBub on 13 Jul 2010 07:58
Bob Adkins wrote:
> On Mon, 12 Jul 2010 09:01:44 -0500, "HeyBub" <heybub(a)NOSPAMgmail.com>
> wrote:
>
>> Hmm. I have NOT been experimenting with either good or horrible
>> registry cleaners for much longer than that and my computer has
>> never been rendered inoperable due to their absence.
>
>
> I use elephant repellent, and it really works. I have never seen an
> elephant around my house, so that proves it works.

Absence of evidence is not evidence of absence.


First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4 5
Prev: DPC Latency Checker
Next: alt.comp.freeware links at Sat Jul 10 21:20:01 2010