• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Does Windows rot over time?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

What causes Windows to die over time?


  • Total voters
    913
Sjaak said:
Do you want to talk about it? How did it make you feel?


Actually I feel good about my opinion,Xp has never failed me and does all i could want it to do with no crashes or errors ever..So basing my opinion on those facts I think my vote was spot on at least in my case :)
 
I voted for network/NIC issues, 3rd party software issues, and failed overclocking. I have without question experienced degraded performance and file system corruption with bad overclocked memory settings. I've also noticed drops in performance after installing certain applications and drivers. And of course, there's no question that spyware/adware will corrode the performance of Windows.
 
Yeah windows rots.

Between but I think it is other stuff that rots it. Viruses, user installed software, ect, adds to the deterioration.
 
i did not select any of the above.

first off, i believe any of the above (weather a poll selection and/or something mentioned my a member above) CAN and frequently DOES cause eventual decay, and has plenty if accidents waiting to happen with a green install, i find that *above all else*,the PEBCAC.

people who operate a computer like a microwave ( just use it, and wipe it with a damp sponge occasionally) are far more likely to completely and apocaliptically destroy every last line of code on the machine, to the point of just booting into safe mode every day, only months or years later whining to someone capable of fixing it. they dont back up anything, and in tears when all their data is gone-when they ask about a "failure proof" computer they are shocked that their $2000 machine has little or no failure-proofing.
sure they ask sometimes - "can i buy one that *wont* break?"
and when quoted a price for a workstation with a big UPS, redundant PSU, ECC ram, and fault tolerant SCSI system they send you pakcing when they hear it costs 3 times as much, and cant understand why their computer has none of the above in the first place, since they paid so much for it allready... even a machine like that wont save them. i would wager more often than not, they could kill it within an hour, as compared with a few weeks with the "sponge method".

ive seen it, and ive seen plenty of it. its a true horror.

there are two lines of thinking about this :
1. it is the fault of microsoft for expecting people to have knowledge of more than a sponge for maintenance.
2. its the fault of the user for not understanding that a computer is incomprehensibly complex, and a wet sponge is the last thing they need.
thats a never ending debate.

it is tied closley to human psycology, that i think software makers fail to understand:

the things is, like a refirgerator, people spend a grand or two, come home with a big box, set it up and expect it to JUST WORK, and be that way for years - with occasional dusting. honestly, *i cant fault them for that.* i want my expensive box to *just work* too...

think about it this way:
what if we needed to maintain cars like we needed to maintain computers?
what if your new set of fuzzy dice suddenly caused your car to crash?
what if a grasshopper smashed on the windshield prevented you from locking the doors?
what if you knocked over your coffee mug, and your car only ran in 1st gear untill you replaced the carpets, and not any old carpets, *certified upholstry* and only that?

my personal belief is that an operating system should not break with the "wet sponge" method- would you spend 10 minuets a day working on your washing machine?

allso worthy of note, there are people in the world who will buy a new car, never wash it, never change the oil, never rotate the tires, hit potholes, clip curbs, and smash up the body panels, day in and day out with a cold one in one hand, a cell phone in the other, a stogie in their mouth and a girly mag in their lap. yes, there are people like this in the world. i bet my life -
fortunately, most of them never own a computer, or those who do we dont have the pleasure to encounter.

the flip side is people like us. sure we know how to maintain things, in addition to accepting blindly that such things *can, do, and should* require such efforts...we can breeze through the registry and clean it up manually. we can find compatable drivers in the dark corners of the internet. we keep out holes patched, and keep the big bad bugs out, we know enough to buy enough memory so we dont live in the swap file, and can even spend 1/4 of out time at the keyboard keeping things shipshape. its a real drag!!
just because we know how to more or less keep things proper does not mean that its the "right way" - okay, so we spend 1/4 of the time in the chair doing housekeeping and not actually USING the machine. *even then* we sometimes still have problems, and still see our fair share of catastrofic failures.

yes, we are at the wheel of an operating system that -by default in case they interfere or get intrusive, has all maintenance features turned off-, and never hinted at, but will *automatically* download and install and run any number of evil pieces of software and run them all the time weather you like it or not, from millions of websites.

so then, am i for a dumbed down operating system that maintains itself and scans itself for big bad bugs, and fixes everything anyway without your knowledge?
yes.
but the flip side of dumbing down requires an "advanced" button, or tab somewhere for it- with an idiot warning, because there are allready too many things inaccessable.
if i get into a situation that requires an "advanced" tab, and its not there, i would be rightfuly upset.

in summation, its a case of blind human psycology meeting blind software code, with no diplomacy between. most often, its the human factors that lead to problems.
less often (though it happens all the time) its something else related to non-human-induced-decay.
 
Educating users won't entirely cure Windows of its problems. Manufacturers like Dell need to learn that XP just doesn't run well in (really) less than 512MByte. XP disk-thrashes frequently in less than 512M.

Even on my main box, with 2GB of RAM, the system swaps. I don't like this. If you've got physical memory available, why would you _deliberately_ swap and bottleneck things? Granted, paging to disk is an advance in technology, but that doesn't mean it should be used constantly.
 
Is there some way to tell when windows was originally installed? I have files from 2003 in my system32 directory, and have never reinstalled.
I *never* get spyware or viruses, and I do not run an anti-virus program in the background. I sometimes run SpyBot, Adaware, and online virus scanners, but they have never found anything more than cookies or MRU lists.
I have always been behind a firewall, since I also need something to share the internet connection with several computers.

Spyware and viruses are the only reason that Windows XP/2003 would "rot" over time;my installations simply keep going and going.

P.S. I'm not just a windows fanboy, BTW. I have debian 3.1 set up as dual-boot on my machine.
 
DaBigJ said:
Is there some way to tell when windows was originally installed? I have files from 2003 in my system32 directory, and have never reinstalled.
If you go into your C drive and right click on your Windows folder and select properties you will see the installation date under date created.
 
Quailane said:
That is really surprising. Windows seems fine after even a few years if you don't let spyware get to it. We have had an athlon xp with windows xp since 2001 and it has not "rotted" or whatever at all. And that is with me and my brother treating it poorly, but keeping spyware off. Second longest is 2 years with another athlon xp. No problems or massive slowdowns at all. The older athlon xp is not overclocked at all and the newer one is all within spec except the cpu. That is with non-motherboard hardware changes with them every once in a while.

I know everyone always bashes microsoft. I don't think they would make an OS that goes bad like that. I have a friend who runs windows 95 on a 95Mhz Pentium. Yeah it feels a little slow because the hardware is a lot slower than today, but it still connects to the internet and is useful for typing if someone else is on the other computer. That is 10 years or so of working fine compared to your 4 months. The only time I ever reinstall is if I change the motherboard (and not even sometimes then) or if windows gets eaten alive by a virus or spyware (only did once). It must be something you are doing.

Good data there :thup:

I have strong circumstantial evidence/suspicion that my NVidia Ethernet + associated drivers has some issue with the OS. My benching drive uses the Marvell onboard LAN and it seems to be more stable. With the NV LAN on my daily usage drive, I get BSOD's occasionally when downloading large files i.e Linux iso's 3DM etc. Bear in mind that I use my benching drive more than the regular drive :D.

I don't know what it is I'm doing wrong here. I always test an OC with the benching /test install before moving it to regular usage? Just compiled kde this morning. Not one error or seg fault.

Edit*
Orion,
You give Windows too much credit. Nobody here (atleast no rational person) is hating on Windows. What I or others express is a result of the frustration one faces with such a vexing issue. My attempt here is to facilitate the exchange of information which can minimize such problems. In order to do that, I'm trying to determine (using a broad swathed poll) the constraints one has to work under, to keep things under control. I do appreciate your philosophical point of view, but my intent was to get a practical response, preferably with evidence, anecdotal or otherwise.
 
Last edited:
Windows does indeed rot over time. I have to reformat and reinstall windows every 1-2 months.

Linux, on the other hand will work for however long without reformatting or even rebooting.
 
Chris_F said:
Windows does indeed rot over time. I have to reformat and reinstall windows every 1-2 months.

Linux, on the other hand will work for however long without reformatting or even rebooting.
To me, this is a key benchmark in OS-coolness. It's unheard of, for me, to have a system stay up for a day under XP (mostly due, admittedly, to drivers and such). My #2 linux box has been up for *clicky click 'uptime'* 6 days 5 hours 2 minutes, and not so much as a moan.

Although, for what it's worth, most desktop users don't need a system to stay up for 6 days.

I've done some more thinking on the topic of Windows and, more generally, operating system integrity, and have come to the conclusion that Windows is meant for systems that aren't meant to stay up that long. Linux (and other Unix derivatives) originated in the big iron days, and are meant to operate servers. Thus, a divergence in philosophies.

Perhaps the day will come when there are two main varieties of OS: Server OSes and Desktop OSes. I mean this in the sense of completely different products, being developed by different groups, with different internal workings, that can exchange data with each other. (This is slowly happening today, I think, with the usage of Linux for many server applications and, invariably, the use of Windows on the destkop. Completely different products, that, in spite of completely different development projects, work together.)
 
Mate (Chris),

Is that an objective opinion? Could you add a bit more regarding your experience. Looking at your sig, I can see that you are not an MS happy camper :)

Newbie,
I believe Linux is all about choices (atleast Debian, BSD and Gentoo are). Whereas with Windows one has no choice. I installed a fully optimized XP using nlite. Removed quite a bit of the fat (unwanted drivers..etc). I keep getting a nag screen saying that some files were not original windows files and I had to replace them. This clearly shows that MS are more worried about people customizing their systems than with producing a quality product.
 
Last edited:
Windows is very susceptible to viruses and spyware. The windows registry is susceptible to getting ****ed by viruses and when you install and uninstall a lot of software. Before you know it windows takes 5x longer to load then it did before, task manager shows you have 200 processes running and you can't even open a program without the computer bluescreening. :p

It's the same story every time.

If you know what your doing you can try your best to maintain windows but you usually cave in and decide to wipe the dang thing and reinstall it.
 
I believe Linux is all about choices (atleast Debian, BSD and Gentoo are). Whereas with Windows one has no choice. I installed a fully optimized XP using nlite. Removed quite a bit of the fat (unwanted drivers..etc). I keep getting a nag screen saying that some files were not original windows files and I had to replace them. This clearly shows that MS are more worried about people customizing their systems than with producing a quality product.
Also agree. The degree of power and freedom of configuration in Linux, however, resulted from its hackish heritage ("You _can_ change this if you absolutely want to, we'll give you that freedom"), while the lack of configuration in Windows is directly attributable to its corporate and quote-unquote professional development heritage.

I'm trying to take an objective look at this.

They (Windows and Linux) are admittedly developed by different methods for different reasons.
Linus Torvalds' Just For Fun said:
Originally, this was a project to teach me about the 386.
This continues today. A lot of linux ports exist; Windows doesn't support a lot of hardware, for reasons that I'll pontificate on below.

Additionally, from the start, a willingness to share Linux was demonstrated. Although Richard Stallman would disagree, I don't think the fact that Linux was based on gnu tools has much to do with it. Instead, I think it is simply an extension of the Unix culture to the desktop. This culture is based on the days when only academic institutions had computers.

In the Olden Days, you shared what worked, and even shared what didn't, so that the whole world could learn about it, and those lessons would be reworked and circulated throughout the community. To this day, that philosophy is still embodied in Linux and BSD. This is owed to the fact that academic institutions tend to be very open. It's very easy to get documentation on Linux's/BSD's APIs.

Windows, on the other hand, was developed/is developed under tight secrecy in a corporate environment, for very specific purposes. The IBM PC, to which most personal computers owe its heritage, was admittedly far inferior to the big boxes of the time. It had no notion of multitasking or multiusership. It had no notion of networking. It had some features that, while it made the PC and subsequent machines affordable, didn't help much in terms of making the PC a real computer in the sense of multiuser networked PCs.

And that's okay. The original PC wasn't intended to have any of those features. It was meant for you to write documents and play Reader Rabbit on. The introduction of PC clones meant that IBM/Microsoft (when they were actually working with one another) had to start looking at supporting a _lot_ of hardware, and supporting it all uniformly. DOS was never meant to run more than one program at a time. It wasn't ever meant to have more than 1MB of RAM. Windows 1.0 added a graphical user interface, making the PC more of a System for the Rest of Us. Because of the lack of protected mode, cooperative multitasking was used for the Win16 versions (1-3.11, 95, 98, Me, to a lesser extent in the newer versions). A lot of the admittedly inherent instability in later Windows versions is owed to the fact that there is still that lurking layer of cooperative multitasking up to and until Windows XP.

(NT never had any of these problems. NT was developed for different purposes--a serverside operating system and an operating system designed for higher demands. It was meant to be 32-bit right from the start.)

Then there were the technical advances from Intel--the 286, then the 386, and the 486--which forced Microsoft to add support for modern concepts such as virtual memory, protected mode, and 32-bit addressing, and preemptive multitasking.

Even Windows ME, although a 32-bit hack on a 16-bit Windows system, still has that DOS layer lurking underneath.

And then there was the Internet. By this time, almost every household that had a computer had an IBM PC or compatible. The simple fact that Microsoft was firmly established already set the standard. And suddenly it was important for the PC to have security functions. The lack of a user security model made this even more difficult.

Windows and DOS became layer upon layer of patches, hacks, and extensions. Temporary fixes, in all reality, became permanent, and stayed around because it's somewhat important and good for marketshare to not force anyone to get new software when you release a new OS.

The golden handcuffs of compatibility with those features remain in Microsoft Windows to this day (although, after Millenium, the last vestiges of Win32-over-DOS were eliminated). It's still possible to run Windows 3.1 code on XP, with a little prodding, but no Win16 application is able to sieze control of the system--XP is the boss on Windows systems, now and forevermore, since boot. DOS is still there, but it's no longer in charge: it runs as a virtual machine.

The NT technologies coming to the consumer version of Windows, through WinXP, has given me a lot of new respect for the Windows platform.

And, although x86 with Windows is the most common platform/OS on the planet, there are two distinct types of developers for it: those that don't have a whole lot of professional training from Microsoft about the idiosyncrasies of their OS, and professional engineers, educated by Microsoft.

Finally...Windows will run on all x86 platforms, but isn't ported to anything else. Suport for SPARC, MIPS, and PowerPC was axed after NT4. This is simply because the majority of the world runs on x86 platforms, even serverside, and it wasn't economically sound to pay the PowerPC/... units of the NT developer corps for code that wasn't used a lot. And since the sources weren't open, PowerPC NT4 died.

Just a whole bunch of musings. (This, I think, is worth a little more than two cents. Call it a nickel. :) )
 
Voted 2, 3, and 4.

Never ever had a problem that was caused by Windows by itself... always user error. Be it OCing, tweaking, or downloading that little flash game... ALWAYS the user. People who download, install, uninstall, play, click there, click here, use IE, click banners, keep AV out of date... are just begging for issues. If you're smart, it's completely possible to have a completely secure, functional XP system for months or years. I know, I'm one of those people. Only time I have to reformat is when I do stupid stuff... 'nuff said.
 
IMHO it's usually b/c the user is stupid and downloads stuff filled with spyware and viruses, or maybe they just mess up a little tweaking it, and because of one little file being deleted, the entire OS stops working. That's why I like unix based systems. Due to the different file architecture and the relatively low number of users (which means there aren't many people to program spyware and viruses for them), unix systems are practically immune to spyware and viruses. But you have to have an IQ of like 200 to use it right :p
 
I can attest to the fact that windows becomes more bloated and slow over time - on this 30-week-old-install the registry has been growing by 2MB per month and getting more sluggish. The plethora of applications I have installed all want to input explorer shell extensions, protocol handlers and other stuff to nicely bog down my system even further. Of course, deleting random windows installer executeables from C: was probably not a good idea but I have managed to keep C:\Windows under 1.5GB for an XP Professional SP2 install.
 
Back