Why windows xp becomes very slow after a few months-a year?

tryon wrote:

> I''m taking the operating system class this semester and a question
> was brought to my attention: Why windows xp becomes very slow after
> a few months-a year?
>
> Let's assume here, for the sake of the discussion, that this is the
> case for 90%+ of the system, excluding those in the business world
> where users are restricted from installing everything they want,
> and even there...
>
> My initial thoughts are that it may be related to the way the OS
> handles the registry which can become very messy quickly. (As you
> can see in gates's internal memosl )
> The other one might be initial flaw in the NTFS and the way xp
> handles it. (Which doesn't make much sense since as far as I know
> windows vista doesn't become as slow as xp after a few months and
> they both use NTFS... even though I know some might say "vista is
> running at the speed of xp after 6 months, at the beginning")
>
> I'm only beginning to learn the in-depth insides out of an OS but
> I'm sure you guys have some more in-depth knowledge of xp os versus
> alternatives (as far as I know there's no such thing as the "format
> and reinstall" problem solving technique in OS X or gnu/linux)
>
> I've started reading Operating System Concepts 7th edition by
> Silberschatz.Galvin (with the dinosaurs on the cover) but I doubt
> that this will explain the flaw in the way a system was implemented
> versus another choice and that's what really interest me. If you
> have suggestions of links/video/webcast/charts on the
> implementation of the windows OS it would be appreciated
>
> I'd like the comments to stay on the technical reason for this
> almost-constant decrease in performance in XP. I'm looking forward
> for your answers and I'll ask my teachers this week

I've been using this install of Windows XP (without a wipe) since early
2002. I have changed hardware on it, installed and uninstalled likely
hundreds of pieces of software, etc - and now it is a virtual machine. It
did not 'slow down' (in an unexpected manner - to be explained next in this
response) in all that time.

Sure - if I installed a newer application on top of Windows XP that required
more resources than its previous version (Microsoft Office 2000 to Microsoft
Office XP to Microsoft Office 2003 to Microsoft Office 2007 or Adobe
Photoshop 6 to Adobe Photoshop 7 to Adobe Photoshop CS# and so on...) -
those applications seemed slower than their previous counterparts - because
they were on my system as it stood - which had not changed in available
resources even though I then demanded more from it.

A Windows XP machine - in my experience with my own 1/2 dozen or so machines
and the thousands that I manage (some of which fit your 'the users cannot
install software' description) or assist with - does not slow down over
time/usage. Perception might be that it is slower if the user does get to
use newer/faster technology and/or they upgrade their software with no
upgrade to their hardware to assist with the new demands... But as far as
it actually being slower - I have not seen it.

It's all a matter of proper system maintenance. If a machine is maintained
in a manner that keeps it from getting infested/infected without overloading
it with applications that help when common sense is lacking - I find the
machine can go at the same speed it originally performed at for years.

Most perceived slowdowns can usually be attributed *not* to the operating
system (although there may be _some_ as service packs/patches make the code
more complicated/throw in more security checks) but to the applications
being installed and utilized on top of the operating system. People seem to
forget that they are not running pure Windows XP when they open Internet
Explorer and visit a web page that has Java, Flash, Shockwave, etc all over
it or when they open up the office suite of their choice and compose a
document, work on a spreadsheet, create a multimedia presentation... Very
few people actually just use the operating system and never install anything
on top of it.

The system I mentioned in the beginning and now use as a virtual machine has
gone through so many software installs/uninstalls that the registry is
overflowing with entries that have *no* purpose - yet I have had no need
(based on the performance of the system) even using it daily - to try and
clean it up. I have - as should be obvious from above - given it the
hardware upgrades it needed over the years to compensate for the new demands
I would put on it with new software installed on top of the operating
system.

So - therefore - I feel a discussion over the statement that a Windows XP
system "... becomes very slow after a few months-a year ..." is going to be
a very short one. Windows XP is - no doubt - running as fast as it was the
day it was installed (barring hardware failures.) At worst - it *might* be
running slightly slower because of a patch that complicated something due to
security checks, etc - but then again - the patch may have made the code
more efficient - speeding up performance (albeit - overall - slight and
likely not noticeable.)

Now - a computer seldom sits idle for a few months/a year - seldom can you
come back to a system after that amount of time and find *everything* the
same version as you left it... Nothing new installed at all. So what makes
the conversation short - what points out the direct flaw in the logic - is
that the operating system is unlikely the culprit in the slowness (if any)
being perceived because it is not alone in the picture, by far.

Try to use a pure Windows XP system for a few months/a year. See if you get
frustrated not being able to go to your favorite web page, open that PDF,
work on that document and check the spelling, work on that spreadsheet and
even chat with friends over IM (the old version of Messenger installed isn't
going to be much help.) Most likely - if you avoid getting
infected/infested (you can - that just requires common sense and the fact
you have none of the plug-ins that might lead you to web pages that could
infest you) - at the end of the few months/a year, you will be able to test
and see that Windows system is just as fast as it was the day you started
using it.

Now - you could argue that this is not how people use their machines... Aha!
You are correct. No doubt. And without a doubt - the majority of those
users do not perform the periodic maintenance that common sense and best
practices that can be found everywhere with a simple search of the Internet
dictate.

A piece of machinery of any type will run better and for a greater length of
time when it is properly maintained. Take a door lock as an example. If I
keep shoving different (wrong, perhaps) keys in it over and over, heck -
sometimes not even keys - but knives and picks and sticks and twigs as
well... I bet that lock does not last as long as the one I used only the
correct key when I needed to and lubricated periodically.

So - although I know people get annoyed that some people always blame the
end-user for things on their computer (slow, infestations, etc) - this is
one of those things that with some research, knowledge and common sense -
the end-user has a large amount of potential influence on the outcome.

They can determine if the machine is properly maintained. They can
determine if programs are installed and uninstalled as cleanly as possible.
They can make judgment calls on whether upgrading to Adobe Photoshop CS4 (or
Microsoft Office XP) will require a hardware upgrade (or two) to assist with
the extra resources needed over their current installation of Adobe
Photoshop 6 (or Microsoft Office 2007.) They can research to determine
which protection/security applications they really need and which ones use
the least system resources at all times. They can ensure their machine's
software (operating system and everything installed on top of it) stays
current/patched and then reflect on/decide upon the resources needed with
any updates/upgrades they do to the software. They can determine which web
pages they visit, when they click on something, when they run something,
when they install something new...

*shrug*

In the end - I disagree with the premise/presentation of the question
entirely. Even if interpolated/interpreted as a whole situation (more than
just Windows XP) - even within realms of reasonable expectations - I see the
query/assumed conclusion as a flawed one, at best.

A properly maintained Windows XP system (and the software installed upon it)
will continue to run fine (at the same level of performance) for a long
period of time (long being at least 6 to 7 years at this point.) Sure - you
might have to upgrade some hardware - but you'd have to do that with any
device/technology you started demanding more of than the original
configuration was designed for. Sure, you might have to understand some of
the inner-workings (ever-so-slightly), so to speak, to make sure you
properly maintain this piece of equipment (your Windows XP system.) Welcome
to the reality of a quickly changing system. Windows XP (base) has stayed
somewhat the same (although when you consider the number of changes even it
has gone through since its release at the end of 2001/beginning of 2002 -
it's not really the same) - but all the stuff people pile on top of it to do
their 'daily work/play'; those are not even close to being the same and some
of them did not even exist.

But, hey, that's just *my* take on it. ;-)

--
Shenan Stanley
MS-MVP
--
How To Ask Questions The Smart Way
http://www.catb.org/~esr/faqs/smart-questions.html

Entire Conversation (Archived Indefinitely):
http://groups.google.com/group/micro...ba504256218bd5



Shenan Stanley wrote:
<snip for corrective action>

> They can make judgment calls on whether upgrading to
> Adobe Photoshop CS4 (or Microsoft Office XP) will require a
> hardware upgrade (or two) to assist with the extra resources needed
> over their current installation of Adobe Photoshop 6 (or Microsoft
> Office 2007.)

<snip for corrective action>

Above I made a mistake. The line should be:

"They can make judgment calls on whether upgrading to Adobe Photoshop CS4
(or Microsoft Office 2007) will require a hardware upgrade (or two) to
assist with the extra resources needed over their current installation of
Adobe Photoshop 6 (or Microsoft Office XP.)"

I somehow swapped the Office Versions - I had someone performing a
downgrade. My bad!

--
Shenan Stanley
MS-MVP
--
How To Ask Questions The Smart Way
http://www.catb.org/~esr/faqs/smart-questions.html

On Mon, 19 Jan 2009 13:41:01 -0800, tryon
<tryon@discussions.microsoft.com> wrote:

> I''m taking the operating system class this semester and a question was
> brought to my attention: Why windows xp becomes very slow after a few
> months-a year?
>
> Let's assume here, for the sake of the discussion, that this is the case for
> 90%+ of the system, excluding those in the business world where users are
> restricted from installing everything they want, and even there...
>
> My initial thoughts are that it may be related to the way the OS handles the
> registry which can become very messy quickly. (As you can see in gates's
> internal memosl )
> The other one might be initial flaw in the NTFS and the way xp handles it.
> (Which doesn't make much sense since as far as I know windows vista doesn't
> become as slow as xp after a few months and they both use NTFS... even though
> I know some might say "vista is running at the speed of xp after 6 months, at
> the beginning")
>
> I'm only beginning to learn the in-depth insides out of an OS but I'm sure
> you guys have some more in-depth knowledge of xp os versus alternatives (as
> far as I know there's no such thing as the "format and reinstall" problem
> solving technique in OS X or gnu/linux)
>
> I've started reading Operating System Concepts 7th edition by
> Silberschatz.Galvin (with the dinosaurs on the cover) but I doubt that this
> will explain the flaw in the way a system was implemented versus another
> choice and that's what really interest me. If you have suggestions of
> links/video/webcast/charts on the implementation of the windows OS it would
> be appreciated
>
> I'd like the comments to stay on the technical reason for this
> almost-constant decrease in performance in XP. I'm looking forward for your
> answers and I'll ask my teachers this week Smiley

You say "almost-constant decrease in performance in XP," but that's
not been my experience at all--not on any of my machines that run XP
now, that ran it in the past, nor on any others that I know.

For those machines where you do see it, I believe the most common
reason is malware infection.

--
Ken Blake, Microsoft MVP - Windows Desktop Experience
Please Reply to the Newsgroup

tryon said this on 1/19/2009 4:41 PM:

> I''m taking the operating system class this semester and a question was
> brought to my attention: Why windows xp becomes very slow after a few
> months-a year?
>
> Let's assume here, for the sake of the discussion, that this is the case for
> 90%+ of the system, excluding those in the business world where users are
> restricted from installing everything they want, and even there...
>
> My initial thoughts are that it may be related to the way the OS handles the
> registry which can become very messy quickly. (As you can see in gates's
> internal memosl )
> The other one might be initial flaw in the NTFS and the way xp handles it.
> (Which doesn't make much sense since as far as I know windows vista doesn't
> become as slow as xp after a few months and they both use NTFS... even though
> I know some might say "vista is running at the speed of xp after 6 months, at
> the beginning")
>
> I'm only beginning to learn the in-depth insides out of an OS but I'm sure
> you guys have some more in-depth knowledge of xp os versus alternatives (as
> far as I know there's no such thing as the "format and reinstall" problem
> solving technique in OS X or gnu/linux)
>
> I've started reading Operating System Concepts 7th edition by
> Silberschatz.Galvin (with the dinosaurs on the cover) but I doubt that this
> will explain the flaw in the way a system was implemented versus another
> choice and that's what really interest me. If you have suggestions of
> links/video/webcast/charts on the implementation of the windows OS it would
> be appreciated
>
> I'd like the comments to stay on the technical reason for this
> almost-constant decrease in performance in XP. I'm looking forward for your
> answers and I'll ask my teachers this week Smiley

I've been running the same install for a year now and my other machine
is closer to 3 years. Neither seem to be slowing down. I will
admit that I get more impatient and thus the PC seems to slow down, but
in reality I can't measure the machine slowing down. I agree Shenan
in that I think Photoshop CS4 is so slow on my PC compared to Elements
2.0. And that is expected since CS4 is so much more complex and
bloated over E2.

Big_Al <BigAl@md.com> wrote:

>tryon said this on 1/19/2009 4:41 PM:

>> I''m taking the operating system class this semester and a question was
>> brought to my attention: Why windows xp becomes very slow after a few
>> months-a year?
>>
>> Let's assume here, for the sake of the discussion, that this is the case for
>> 90%+ of the system, excluding those in the business world where users are
>> restricted from installing everything they want, and even there...
>>
>> My initial thoughts are that it may be related to the way the OS handles the
>> registry which can become very messy quickly. (As you can see in gates's
>> internal memosl )
>> The other one might be initial flaw in the NTFS and the way xp handles it.
>> (Which doesn't make much sense since as far as I know windows vista doesn't
>> become as slow as xp after a few months and they both use NTFS... even though
>> I know some might say "vista is running at the speed of xp after 6 months, at
>> the beginning")
>>
>> I'm only beginning to learn the in-depth insides out of an OS but I'm sure
>> you guys have some more in-depth knowledge of xp os versus alternatives (as
>> far as I know there's no such thing as the "format and reinstall" problem
>> solving technique in OS X or gnu/linux)
>>
>> I've started reading Operating System Concepts 7th edition by
>> Silberschatz.Galvin (with the dinosaurs on the cover) but I doubt that this
>> will explain the flaw in the way a system was implemented versus another
>> choice and that's what really interest me. If you have suggestions of
>> links/video/webcast/charts on the implementation of the windows OS it would
>> be appreciated
>>
>> I'd like the comments to stay on the technical reason for this
>> almost-constant decrease in performance in XP. I'm looking forward for your
>> answers and I'll ask my teachers this week Smiley

>
>I've been running the same install for a year now and my other machine
>is closer to 3 years. Neither seem to be slowing down. I will
>admit that I get more impatient and thus the PC seems to slow down, but
>in reality I can't measure the machine slowing down. I agree Shenan
>in that I think Photoshop CS4 is so slow on my PC compared to Elements
>2.0. And that is expected since CS4 is so much more complex and
>bloated over E2.

And there are NUMEROUS complaints about its slowness on the Adobe
board.

I'll stay with CS3 ;-)

"Shenan Stanley" wrote:

>
> I've been using this install of Windows XP (without a wipe) since early
> 2002. I have changed hardware on it, installed and uninstalled likely
> hundreds of pieces of software, etc - and now it is a virtual machine. It
> did not 'slow down' (in an unexpected manner - to be explained next in this
> response) in all that time.
>
> Sure - if I installed a newer application on top of Windows XP that required
> more resources than its previous version (Microsoft Office 2000 to Microsoft
> Office XP to Microsoft Office 2003 to Microsoft Office 2007 or Adobe
> Photoshop 6 to Adobe Photoshop 7 to Adobe Photoshop CS# and so on...) -
> those applications seemed slower than their previous counterparts - because
> they were on my system as it stood - which had not changed in available
> resources even though I then demanded more from it.

Of course some newer apps are going to require more power to let the system
runs as smoothly as before, but that’s only if the apps are started. You gave
a good example with the office suite. Now if you have installed/uninstalled
-which almost always leave some registry/folders « crap » behind- there’s a
lot of things we know the default uninstaller doesn’t clean up very well, you
have to do it on your own and not all the tools necesary to do so are
included in xp, and would you need these tools in the first place, shouldn’t
it be automated?. That’s one of the issue with the way the OS works that
bothers me.

> A Windows XP machine - in my experience with my own 1/2 dozen or so machines
> and the thousands that I manage (some of which fit your 'the users cannot
> install software' description) or assist with - does not slow down over
> time/usage. Perception might be that it is slower if the user does get to
> use newer/faster technology and/or they upgrade their software with no
> upgrade to their hardware to assist with the new demands... But as far as
> it actually being slower - I have not seen it.
>
> It's all a matter of proper system maintenance. If a machine is maintained
> in a manner that keeps it from getting infested/infected without overloading
> it with applications that help when common sense is lacking - I find the
> machine can go at the same speed it originally performed at for years.

75% of geeks prefer manual cars, (the reasonning here is that its give us
more control but requires more work/harder work), but the general population
simply wants something that works. (from Why Software Sucks by David S Platt)

They don’t want to learn how to clean their registry / defragment their HDs
/ uninstall the mess a software leaved behind / tweak their os / launch the
cleanup utility / any other task a geek might consider as « basic system
maintenance ». Around 12-14% of the cars in the US are sold with a stick
shift.
An OS performance shouldn’t rely on a task the user know nothing about and
don’t wanna get involved in. That’s, I think, another basic design flaw.

Why not offer to automaticly schedule such tasks?
Why not defragment the HD when it’s idle ? (as some 3rd party vendor offer)
Why consider the user responsible for performing OS maintenance tasks
because of the choices that were made when it twas design/created?

> Most perceived slowdowns can usually be attributed *not* to the operating
> system (although there may be _some_ as service packs/patches make the code
> more complicated/throw in more security checks) but to the applications
> being installed and utilized on top of the operating system. People seem to
> forget that they are not running pure Windows XP when they open Internet
> Explorer and visit a web page that has Java, Flash, Shockwave, etc all over
> it or when they open up the office suite of their choice and compose a
> document, work on a spreadsheet, create a multimedia presentation... Very
> few people actually just use the operating system and never install anything
> on top of it.

Of course if you install windows XP by itself without any additionnal
peripheral/driver/program and let it run by itself for 10 years it will
probably have no problem what so ever during this period of time.
The fact is that’s not what an operating system is for and it’s too simple
to simply blame it all on the end user (I recognize that they deserve part of
the blame though, as Bruce Schneier said : « The user's going to pick
dancing pigs over security every time »). I know that device drivers and 3rd
party programs are not in microsoft’s control but they still managed to
correct some of these concerns in their recent OS even if they don’t have
direct control over them. Providing a uniform model that doesn’t allow a
driver to crash or slow down a whole system -as often as it used to- is a
good start, which was only introduce in vista as far as I know.

> The system I mentioned in the beginning and now use as a virtual machine has
> gone through so many software installs/uninstalls that the registry is
> overflowing with entries that have *no* purpose - yet I have had no need
> (based on the performance of the system) even using it daily - to try and
> clean it up. I have - as should be obvious from above - given it the
> hardware upgrades it needed over the years to compensate for the new demands
> I would put on it with new software installed on top of the operating
> system.

I’m glad to hear that didn’t cause you any issue, but it is a well-known
fact that the registry becomes unusable very quickly as stated by mr. Gates
when refering to the « uninstall » section of windows xp : « Someone decided
to trash the one part of Windows that was usable? The file system is no
longer usable. The registry is not usable. This program listing was one sane
place but now it is all crapped up» (see his 2003 memo to Jim Allchin)

Why let such a thing happen in the first place ?
Why allow something so important for the inner working of this OS becomes
unusable ?
When in the design of the OS this choice happened and why ?

> So - therefore - I feel a discussion over the statement that a Windows XP
> system "... becomes very slow after a few months-a year ..." is going to be
> a very short one. Windows XP is - no doubt - running as fast as it was the
> day it was installed (barring hardware failures.) At worst - it *might* be
> running slightly slower because of a patch that complicated something due to
> security checks, etc - but then again - the patch may have made the code
> more efficient - speeding up performance (albeit - overall - slight and
> likely not noticeable.)
>
> Now - a computer seldom sits idle for a few months/a year - seldom can you
> come back to a system after that amount of time and find *everything* the
> same version as you left it... Nothing new installed at all. So what makes
> the conversation short - what points out the direct flaw in the logic - is
> that the operating system is unlikely the culprit in the slowness (if any)
> being perceived because it is not alone in the picture, by far.
>
> Try to use a pure Windows XP system for a few months/a year. See if you get
> frustrated not being able to go to your favorite web page, open that PDF,
> work on that document and check the spelling, work on that spreadsheet and
> even chat with friends over IM (the old version of Messenger installed isn't
> going to be much help.) Most likely - if you avoid getting
> infected/infested (you can - that just requires common sense and the fact
> you have none of the plug-ins that might lead you to web pages that could
> infest you) - at the end of the few months/a year, you will be able to test
> and see that Windows system is just as fast as it was the day you started
> using it.

An OS is not to be used alone so it’s the ecosystem around the OS and the
way the OS handles it that matters to me. In this case I think the system
maintenance necessary when installing/uninstalling lots of software shouldn’t
have to be understood or even done by the end-user in a manual manner, it
doesn’t make sense at an usability stand point.
What part of the OS let this lack of maintenance becomes a problem in the
first place? I believe the XP UI (especially the way to run maintenance
tasks) has very much to do with it.

> Now - you could argue that this is not how people use their machines... Aha!
> You are correct. No doubt. And without a doubt - the majority of those
> users do not perform the periodic maintenance that common sense and best
> practices that can be found everywhere with a simple search of the Internet
> dictate.

You are refering to common sense, I’m sorry but what we, so called computer
profesionnals, called « intuitive » isn’t for most people, becuase they
haven’t learned what a plugin or an active x is and they shouldn’t need to in
order to be safe on their computer.(part of the UI, especially dialog box are
responsible for that fact and also the lack of sandboxing, I think) The fact
that the webrowser is so deeply inside the OS creates multiple possible
cross-over that, I think, shouldn’t have occur in the first place and the
user should always have the decision to let the browser interacts with the OS
or not.

> A piece of machinery of any type will run better and for a greater length of
> time when it is properly maintained. Take a door lock as an example. If I
> keep shoving different (wrong, perhaps) keys in it over and over, heck -
> sometimes not even keys - but knives and picks and sticks and twigs as
> well... I bet that lock does not last as long as the one I used only the
> correct key when I needed to and lubricated periodically.

Of course and I agree with the fact that the more you maintain something,
the longer it will stay fit, but the user isn’t interested in maintaining
their os, they wanna use softwares. The maintenance tasks should be mostly
automatic and easy to reach. (sorry but hiding the clean up utility isn’t
really bright. It could have been done the same way than when the OS ask us
if we want to remove old icons from the desktop. E.g., it might have asked us
if we wanted to defragment/clean up the disk once in a while) That’s for me
another major design flaw in usability, but not in the inner workings of the
OS itself which is what I’m looking for.

> So - although I know people get annoyed that some people always blame the
> end-user for things on their computer (slow, infestations, etc) - this is
> one of those things that with some research, knowledge and common sense -
> the end-user has a large amount of potential influence on the outcome.
>
> They can determine if the machine is properly maintained. They can
> determine if programs are installed and uninstalled as cleanly as possible.
> They can make judgment calls on whether upgrading to Adobe Photoshop CS4 (or
> Microsoft Office XP) will require a hardware upgrade (or two) to assist with
> the extra resources needed over their current installation of Adobe
> Photoshop 6 (or Microsoft Office 2007.) They can research to determine
> which protection/security applications they really need and which ones use
> the least system resources at all times. They can ensure their machine's
> software (operating system and everything installed on top of it) stays
> current/patched and then reflect on/decide upon the resources needed with
> any updates/upgrades they do to the software. They can determine which web
> pages they visit, when they click on something, when they run something,
> when they install something new...

They don’t want to and shouldn’t have to do those things as this is almost
all related to the way the OS is build and the end-user shouldn’t be
concerned with such details. For example most unix-like OS don’t require
defragmentation as often as fat/ntfs and the user shouldn’t have to be aware
of this particular design choice in NTFS.

> *shrug*
>
> In the end - I disagree with the premise/presentation of the question
> entirely. Even if interpolated/interpreted as a whole situation (more than
> just Windows XP) - even within realms of reasonable expectations - I see the
> query/assumed conclusion as a flawed one, at best.
>
> A properly maintained Windows XP system (and the software installed upon it)
> will continue to run fine (at the same level of performance) for a long
> period of time (long being at least 6 to 7 years at this point.) Sure - you
> might have to upgrade some hardware - but you'd have to do that with any
> device/technology you started demanding more of than the original
> configuration was designed for. Sure, you might have to understand some of
> the inner-workings (ever-so-slightly), so to speak, to make sure you
> properly maintain this piece of equipment (your Windows XP system.) Welcome
> to the reality of a quickly changing system. Windows XP (base) has stayed
> somewhat the same (although when you consider the number of changes even it
> has gone through since its release at the end of 2001/beginning of 2002 -
> it's not really the same) - but all the stuff people pile on top of it to do
> their 'daily work/play'; those are not even close to being the same and some
> of them did not even exist.
>

Ok so you think it’s ok if the usesr « have to understand some of the
inner-workings (ever-so-slightly), so to speak, to make sure [they] properly
maintain this piece of equipment ([their] Windows XP system.) » ?
That wouldn’t be a basic design flaw in your eyes, forcing your using to
understand the way the os is build?

>"Ken Blake, MVP" wrote:
> For those machines where you do see it, I believe the most common
> reason is malware infection.
>
> --
> Ken Blake, Microsoft MVP - Windows Desktop Experience
> Please Reply to the Newsgroup

The user doesn’t know it and doesn’t care, once again. Most will even try to
uninstall malwares with microsoft malware removal tools, spybot or adaware,
but most « good » malware will left the system in a state unusable once it’s
removed as they bury themself deep inside windows. Most of the time, the user
come to us with an unusable system and we have to reinstall it all
over.(sometimes due to this lack of maintenance that shouldn’t have been his
job in the first place) An OS that let the user install programs that might
« kill the OS » so easily is for me a flaw in itself as there’s no way to
remove everything this apps has installed simply by removing it which is what
an end user would expect to be able to do.

Malware aside, another usability issue with the OS (and IE) would be when
installing a suite of applications it sometimes install a toolbar almost
automaticly (you have to notice the checkbox at the bottom) and IE will never
ask you if you really want to install it or not. I very often come around
computer that have 2 to 3 of these toolsbars and the user isn’t eve naware of
it and never knowingly allowed them into their system, so I think it’s unfair
to blame them for everything, the OS should have some protection feature to
protect the users from things running on the computer. (that has partially
been resolved since, with UAC and the whole do-not-run-apps-as-an-admin
issue, but it was still present in xp)

Users don’t want XP, they want to get the job done by using software.
Users don’t care about their OS the way you seem to think they should and
they don’t because they got lifes to live.
I’m wondering what part of the OS itself (not the end-user part) is
responsible for most of the problems when using the softwares. Most people
aren’t geek and won’t maintain their PC themself, that’s a problem we all
have to face as a community (yep I’m using microsoft product too you know)
and all I wanted to know is what was done wrong (or what might have done
better) in XP in term of how the OS was structured inside-out. I know that
asking people – and MVP so it seems- on a microsoft discussion group isn’t
probably the best idea ever but still I wanted your profesionnal point of
view on the inner working parts of the OS and not the discussion on what the
users are doing wrong and the discussions that you never had any issue with
your OS.

<snip>

tryon wrote:

> Of course some newer apps are going to require more power to let
> the system runs as smoothly as before, but that's only if the apps
> are started. You gave a good example with the office suite. Now if
> you have installed/uninstalled -which almost always leave some
> registry/folders « crap » behind- there's a lot of things we know
> the default uninstaller doesn't clean up very well, you have to do
> it on your own and not all the tools necesary to do so are included
> in xp, and would you need these tools in the first place, shouldn't
> it be automated?. That's one of the issue with the way the OS works
> that bothers me.
>
> 75% of geeks prefer manual cars, (the reasonning here is that its
> give us more control but requires more work/harder work), but the
> general population simply wants something that works. (from Why
> Software Sucks by David S Platt)
>
> They don't want to learn how to clean their registry / defragment
> their HDs / uninstall the mess a software leaved behind / tweak
> their os / launch the cleanup utility / any other task a geek might
> consider as « basic system maintenance ». Around 12-14% of the cars
> in the US are sold with a stick shift.
> An OS performance shouldn't rely on a task the user know nothing
> about and don't wanna get involved in. That's, I think, another
> basic design flaw.
>
> Why not offer to automaticly schedule such tasks?
> Why not defragment the HD when it's idle ? (as some 3rd party
> vendor offer) Why consider the user responsible for performing OS
> maintenance tasks because of the choices that were made when it
> twas design/created?
>
> Of course if you install windows XP by itself without any
> additionnal peripheral/driver/program and let it run by itself for
> 10 years it will probably have no problem what so ever during this
> period of time.
> The fact is that's not what an operating system is for and it's too
> simple to simply blame it all on the end user (I recognize that
> they deserve part of the blame though, as Bruce Schneier said : «
> The user's going to pick dancing pigs over security every time »).
> I know that device drivers and 3rd party programs are not in
> microsoft's control but they still managed to correct some of these
> concerns in their recent OS even if they don't have direct control
> over them. Providing a uniform model that doesn't allow a driver to
> crash or slow down a whole system -as often as it used to- is a
> good start, which was only introduce in vista as far as I know.
>
> I'm glad to hear that didn't cause you any issue, but it is a
> well-known fact that the registry becomes unusable very quickly as
> stated by mr. Gates when refering to the « uninstall » section of
> windows xp : « Someone decided to trash the one part of Windows
> that was usable? The file system is no longer usable. The registry
> is not usable. This program listing was one sane place but now it
> is all crapped up» (see his 2003 memo to Jim Allchin)
>
> Why let such a thing happen in the first place ?
> Why allow something so important for the inner working of this OS
> becomes unusable ?
> When in the design of the OS this choice happened and why ?
>
> An OS is not to be used alone so it's the ecosystem around the OS
> and the way the OS handles it that matters to me. In this case I
> think the system maintenance necessary when installing/uninstalling
> lots of software shouldn't have to be understood or even done by
> the end-user in a manual manner, it doesn't make sense at an
> usability stand point.
> What part of the OS let this lack of maintenance becomes a problem
> in the first place? I believe the XP UI (especially the way to run
> maintenance tasks) has very much to do with it.
>
> You are refering to common sense, I'm sorry but what we, so called
> computer profesionnals, called « intuitive » isn't for most people,
> becuase they haven't learned what a plugin or an active x is and
> they shouldn't need to in order to be safe on their computer.(part
> of the UI, especially dialog box are responsible for that fact and
> also the lack of sandboxing, I think) The fact that the webrowser
> is so deeply inside the OS creates multiple possible cross-over
> that, I think, shouldn't have occur in the first place and the user
> should always have the decision to let the browser interacts with
> the OS or not.
>
> Of course and I agree with the fact that the more you maintain
> something, the longer it will stay fit, but the user isn't
> interested in maintaining their os, they wanna use softwares. The
> maintenance tasks should be mostly automatic and easy to reach.
> (sorry but hiding the clean up utility isn't really bright. It
> could have been done the same way than when the OS ask us if we
> want to remove old icons from the desktop. E.g., it might have
> asked us if we wanted to defragment/clean up the disk once in a
> while) That's for me another major design flaw in usability, but
> not in the inner workings of the OS itself which is what I'm
> looking for.
>
> They don't want to and shouldn't have to do those things as this is
> almost all related to the way the OS is build and the end-user
> shouldn't be concerned with such details. For example most
> unix-like OS don't require defragmentation as often as fat/ntfs and
> the user shouldn't have to be aware of this particular design
> choice in NTFS.
>
> Ok so you think it's ok if the usesr « have to understand some of
> the inner-workings (ever-so-slightly), so to speak, to make sure
> [they] properly maintain this piece of equipment ([their] Windows
> XP system.) » ?
> That wouldn't be a basic design flaw in your eyes, forcing your
> using to understand the way the os is build?
>
> The user doesn't know it and doesn't care, once again. Most will
> even try to uninstall malwares with microsoft malware removal
> tools, spybot or adaware, but most « good » malware will left the
> system in a state unusable once it's removed as they bury themself
> deep inside windows. Most of the time, the user come to us with an
> unusable system and we have to reinstall it all over.(sometimes due
> to this lack of maintenance that shouldn't have been his job in the
> first place) An OS that let the user install programs that might «
> kill the OS » so easily is for me a flaw in itself as there's no
> way to remove everything this apps has installed simply by removing
> it which is what an end user would expect to be able to do.
>
> Malware aside, another usability issue with the OS (and IE) would
> be when installing a suite of applications it sometimes install a
> toolbar almost automaticly (you have to notice the checkbox at the
> bottom) and IE will never ask you if you really want to install it
> or not. I very often come around computer that have 2 to 3 of these
> toolsbars and the user isn't eve naware of it and never knowingly
> allowed them into their system, so I think it's unfair to blame
> them for everything, the OS should have some protection feature to
> protect the users from things running on the computer. (that has
> partially been resolved since, with UAC and the whole
> do-not-run-apps-as-an-admin issue, but it was still present in xp)
>
> Users don't want XP, they want to get the job done by using
> software.
> Users don't care about their OS the way you seem to think they
> should and they don't because they got lifes to live.
> I'm wondering what part of the OS itself (not the end-user part) is
> responsible for most of the problems when using the softwares. Most
> people aren't geek and won't maintain their PC themself, that's a
> problem we all have to face as a community (yep I'm using microsoft
> product too you know) and all I wanted to know is what was done
> wrong (or what might have done better) in XP in term of how the OS
> was structured inside-out. I know that asking people - and MVP so
> it seems- on a microsoft discussion group isn't probably the best
> idea ever but still I wanted your profesionnal point of view on the
> inner working parts of the OS and not the discussion on what the
> users are doing wrong and the discussions that you never had any
> issue with your OS.

In the end - yes - I believe if somone is going to use somehting - they
either have to learn how to maintain it or be willing to pay someone to do
it for them.

Cars - they don't change their own oil, go get themsleves
registered/inspected, get their own tune-ups and clean themselves of the gum
wrappers and other assorted items you let get on the floorboards.

Homes - they don't paint themselves, change their own loacks when sold, dust
themselves, repair the A/C and plumbing, and so on...

Lawns don't mow themselves. The mail doesn't just 'appear. in your mailbox.
Your bills are not jsut all automagically setup to pay for you out of some
magical account where money just appears without you doing anything.

Microsoft can up and die for all I care. I support all sorts of OSes every
day - and everyone of them has their problems and even if I combined the
best of all of them - I would have problems with the new OS. The weak link
is that no two people who might use the system are the same and no two
people will likely (therefore) ever react the same so any automated
features/cleanups/etc I build in could do something said user might not want
or could actually be devestating to them.

You cannot solve this problem with software for *everyone*. I have at least
one customer who can take a computer that somone else has been using with
the same rights as them with identical jobs and although the other person
did their job for over a year without a single issue using the same setup -
this person will have an issue within a week.

Maybe the registry design sucks, maybe the way third party installers use it
as a personal dumping ground causes it to be that way. Whose problem is it
that some apps do not uninstall properly? Take Norton as an example - they
have created a Removal Tool to cleanup after their own apps - if they had
just written in the features of that tool into the uninstall portion of the
original apps - they wouldn't need that removal app, eh?

--
Shenan Stanley
MS-MVP
--
How To Ask Questions The Smart Way
http://www.catb.org/~esr/faqs/smart-questions.html