Wiggling the mouse in Windows 95 made the operating system faster

[Shawn Knight]

Through the looking glass: Adding new hardware or overclocking what you've already got are among the leading ways to squeeze more performance out of your PC but in the earlier days of computing, all one had to do to give their PC a jolt was wiggle the mouse cursor around during an install.

I’m a self-professed fidgeter, especially at the computer. It’s not uncommon to catch me moving the mouse all around with zero intent, even if I’m busy reading a story or watching a video.

I chalk it up to the earlier days of computing when keeping the mouse active seemed to have a tangible benefit on system performance. Turns out, it wasn’t just in our imagination.

According to a running thread over on Stack Exchange, moving the mouse cursor in Windows 95 did indeed speed up performance due to a flaw in the way the OS generates events and the fact that many applications are event driven.

As one user explains:

Windows 95 applications often use asynchronous I/O, that is they ask for some file operation like a copy to be performed and then tell the OS that they can be put to sleep until that operation finishes. By sleeping they allow other applications to run, rather than wasting CPU time endlessly asking if the file operation has completed yet.

For reasons that are not entirely clear, but probably due to performance problems on low end machines, Windows 95 tends to bundle up the messages about I/O completion and doesn't immediately wake up the application to service them. However, it does wake the application for user input, presumably to keep it feeling responsive, and when the application is awake it will handle any pending I/O messages too.

Thus wiggling the mouse causes the application to process I/O messages faster, and install quicker. The effect was quite pronounced; large applications that could take an hour to install could be reduced to 15 minutes with suitable mouse input.

As a youth, I’d keep my mouse cursor active as much as possible – and especially during installs – simply because I didn’t want the screensaver to trigger and bog down system resources. Sure, I could have just set the timeout to longer on the screensaver, but that’s beside the point. As it turns out, my actions were actually helpful, and not for the reason I thought.

Found is a TechSpot feature where we share clever, funny or otherwise interesting stuff from around the web.

Permalink to story.

https://www.techspot.com/news/80956-wiggling-mouse-windows-95-made-operating-system-faster.html

 

[Morris Minor]

Lol I still do it now

 

[VitalyT]

I like one that existed since Windows 95 and all the way to Windows 7...

Hovering mouse over standard menu items resulted in timer quickly slowing down in all applications, and then stopping completely.

b.t.w. It was myself who discovered this bug, and I even reported it through proper Microsoft channels a few years back, but then I lost track of it. Maybe I should re-test it against Windows 10 now

 

[ShagnWagn]

I went straight from DOS/Win 3.11 to Win98SE. I held out because of how much a turd 95 was. I don't recall anything like this in 98SE.

Now it's not that 10 is a turd I am passing it over - it's their blatant privacy issues. Perhaps the FTC will fine them several billion dollars like FB got? Let's make it like 10% the value of the company so they actually open their eyes.

 

[Igrecman]

I went straight from DOS/Win 3.11 to Win98SE. I held out because of how much a turd 95 was. I don't recall anything like this in 98SE.

Now it's not that 10 is a turd I am passing it over - it's their blatant privacy issues. Perhaps the FTC will fine them several billion dollars like FB got? Let's make it like 10% the value of the company so they actually open their eyes.

If privacy is a priority for you, you could try W10Privacy
https://www.winprivacy.de/english-home/

 

[Mugsy]

I remember this "glitch".

Moving/clicking the mouse didn't (actually) "speed" anything up. It basically "woke" Win95 up when it got "too focused" on what it was doing, and clicking/moving the mouse alerted the OS that there were other things it could/needed to do.

 

[ShagnWagn]

If privacy is a priority for you, you could try W10Privacy
https://www.winprivacy.de/english-home/

Thanks for the link. I am not a fan of installing 3rd party software to turn off something that should be disabled by default - or at least an option to disable it. The same thing with the junk "cleaner" software. I may consider it.

 

[Skjorn]

Pretty sure this still works

 

[Markoni35]

I know of a stupid limitation (and bug) in file opening, which was introduced in Windows 3.1, because it was limited to 64k segments. Because it was an old 16-bit OS. The bug limits the number of files you can open, and even produces an error if there are "too many" files selected, just because their names can't fit into the 64k memory buffer. Very stupid bug, which affected all the standard Windows apps. And 3rd party apps that used standard Windows dialogs.

But okay, it was a 16-bit limitation. When Windows 95 was released, which was a 32-bit OS, it was expected that the stupid limitation would disappear. Did it? Nope. The old 16-bit limitation was still there, although the OS could address segments that were 65000 times bigger than in Windows 3.1.

And the bug continued in Windows 98, Windows XP, and Windows Vista. But wait... you're probably wondering if they fixed the bug in Windows 8.1 which was 64-bit by default? The answer is: NO. It still had the same limitation, although the OS was now able to address memory segments that were 281,474,976,710,656 times larger than in 16-bit Windows.

But wait... 64-bit Windows 10 has surely removed the limitation, right? By now you probably know the answer: NO. The limitation, and the file-name bug resulting from it, are still there. I bet in the 24th century, one day Klingons will blow away the latest USS Enterprise because of some 16-bit limitation that will still be buried somewhere in the code.

 

[retsxel]

I went straight from DOS/Win 3.11 to Win98SE. I held out because of how much a turd 95 was.

While Win95 had it's problems, calling it a "turd" would be a gross exaggeration.

I don't recall anything like this in 98SE.

It was there.

Now it's not that 10 is a turd I am passing it over

Oh yes it is. LOL!

it's their blatant privacy issues.

And that's one of the many reasons.

 

[retsxel]

If privacy is a priority for you, you could try W10Privacy
https://www.winprivacy.de/english-home/

This looks compelling. Gonna have to give it a VM try..

 

[trparky]

I went straight from DOS/Win 3.11 to Win98SE. I held out because of how much a turd 95 was. I don't recall anything like this in 98SE.

Now it's not that 10 is a turd I am passing it over - it's their blatant privacy issues. Perhaps the FTC will fine them several billion dollars like FB got? Let's make it like 10% the value of the company so they actually open their eyes.

If privacy is a priority for you, you could try W10Privacy
https://www.winprivacy.de/english-home/

As long as you turn the telemetry to Basic Mode there is absolutely nothing in said data that is even remotely close to being an invasion of your privacy. I myself have looked at the data, there's nothing in it that alarms me to any level.

Full Mode is another thing though, Full Mode is a full-blown privacy violation. Good God, the kind of data that is collected with Full Mode is downright scary. But again, turn it to Basic Mode and everything is good to go.

 

[retsxel]

As long as you turn the telemetry to Basic Mode there is absolutely nothing in said data that is even remotely close to being an invasion of your privacy. I myself have looked at the data, there's nothing in it that alarms me to any level.

Full Mode is another thing though, Full Mode is a full-blown privacy violation. Good God, the kind of data that is collected with Full Mode is downright scary. But again, turn it to Basic Mode and everything is good to go.

I usually take that one step further and delete the "Diagtrack" service using the following command from the CLI with admin privileges;
SC delete diagtrack

After that, diagnostic reporting fully ends without a way for it to restart.

 

[Puiu]

I went straight from DOS/Win 3.11 to Win98SE. I held out because of how much a turd 95 was. I don't recall anything like this in 98SE.

Now it's not that 10 is a turd I am passing it over - it's their blatant privacy issues. Perhaps the FTC will fine them several billion dollars like FB got? Let's make it like 10% the value of the company so they actually open their eyes.

Windows 95 (with all of the updates) was one of the best OS MS made. It revolutionised many things and was relatively stable (unlike windows 98, 98SE, ME, XP and XP SP1). It ran like a champ on my 686.

 

[FF222]

Both the statement and the explanation given in the SO thread are absolute rubbish. The so called quantum, which is the time chunk the scheduler allows an application thread to run before scheduling another thread, was 55 ms, which meant a new thread was rescheduled and any outstanding async I/O operations returned/completed 18 times a second. Which was definitely a higher number than the number of mouse events generated in the same time period for a window.

Which in turn means that even if we take the theory behind the mouse event thing for granted (which I'm not, because I never heard of this, despite being very familiar with the internal working of most Windows systems at that time, even though I might be wrong about this), that alone still wouldn't have caused threads getting rescheduled more often than otherwise.

Also, even if re-schedules would have gotten more frequent (because of the mouse or whatever), that wouldn't have made Windows 95 run faster, but actually slower, because now more time would have been wasted on scheduling tasks, than actual processing. So, the only thing that could have been possibly and apparently gotten "faster" would have been only the foreground application at the cost of making everything else slower - but even that is doubtful, considering all the above and how the SO thread doesn't get anything right about how the system works.

I'm pretty positive that if there's any merit to this story at all, and it wasn't just an illusion that anything got faster when wiggling the mouse, then it wasn't the running the application, but that the application window and the progress bar in it was redrawn more frequently, and thus it seemed to skip more often (but not faster), than otherwise, which could have possibly waked the impression of making things somehow faster in the casual - but obviously mistaken - observer. That I would believe any time. The rest of it: not at all, until I actually see proof for that.

 

[kmo911]

How about remaking win 95. cool wiggling mice mouse.older pces like 486 p xx was fast enough. but limited by a sound of moving when you tryed to wiggle mice in som operations. mice hang and program hang. that was fun too. reset and powerbutton. the sound came from monitor sound card.

 

[Igrecman]

As long as you turn the telemetry to Basic Mode there is absolutely nothing in said data that is even remotely close to being an invasion of your privacy. I myself have looked at the data, there's nothing in it that alarms me to any level.

Full Mode is another thing though, Full Mode is a full-blown privacy violation. Good God, the kind of data that is collected with Full Mode is downright scary. But again, turn it to Basic Mode and everything is good to go.

Yes, however with the telemetry to Basic Mode do the services associated with full telemetry shut off or not? I think not. W10privacy can help to turn off services you don't need. The highlighted choices in yellow and red can have consequences on other services though, careful with those. Creating a restore point is a must before making any change with W10privacy. It does ask you to make one when it opens.

 

[James McKeand]

I attended the Windows XP OEM launch event in Washington DC. The Microsoft presenter that was demoing the install process jokingly said that moving the mouse would make the digital sand in the hour glass move faster, thus speeding up the installation.

I guess he wasn’t completely joking...

 

[retsxel]

Windows 95 (with all of the updates) was one of the best OS MS made. It revolutionized many things and was relatively stable

True! By SRC Win95 was great.

(unlike windows 98, 98SE, ME, XP and XP SP1)

Not so true. Win98SE OOBE was easily the most stable version of the Windows 9x run. However(and I know I'm going to get flamed for this), when properly configured WinME was more stable, even more stable than XP until SP2, which is why I personally stuck with ME until then.

You all may laugh at that statement if you wish, but until you've actually run that OS in a proper state(which admittedly needed several config changes after install), you can take a flying leap.

Yes, however with the telemetry to Basic Mode do the services associated with full telemetry shut off or not? I think not. W10privacy can help to turn off services you don't need. The highlighted choices in yellow and red can have consequences on other services though, careful with those. Creating a restore point is a must before making any change with W10privacy. It does ask you to make one when it opens.

No, they do not, see my post above. There are many utilities that can manage services that will help users effect better personal security. However, deleting the offending services is the only way to ensure they do not run behind your back.

 

[CajunMoses]

Windows was archaic then and is essentially that same OS now, only with more bloated baggage.

 

[cliffordcooley]

Windows was archaic then and is essentially that same OS now, only with more bloated baggage.

If I'm connecting all the dot correctly. You are saying Windows is and has always been archaic. The only difference being Windows now has more baggage. If so you might should lookup the definition to archaic.

Windows 95 was not at all archaic, until XP came out. In fact no release is archaic, until after the newer version is released.

 

[retsxel]

XP came out. In fact no release is archaic, until after the newer version is released.

XP is actually still a useful OS for a lot of things. It was only 2 or 3 years ago that hardware vendors stopped supporting it.

 

[Puiu]

Windows was archaic then and is essentially that same OS now, only with more bloated baggage.

Windows is fairly modern, although it does have its own big share of legacy code. The biggest problem windows has right now is the reliance on NTFS for storage.

 

[trparky]

Windows is fairly modern, although it does have its own big share of legacy code. The biggest problem windows has right now is the reliance on NTFS for storage.

Why? NTFS still a very good file system, the fact that it's still around decades after the initial release is a testament to just how good and reliable it really is. Sure, it's undergone some changes and additions but what file system hasn't?

 

[retsxel]

Windows is fairly modern, although it does have its own big share of legacy code. The biggest problem windows has right now is the reliance on NTFS for storage.

NTFS is one of the most stable and well known journaling file systems in the world. There is no need to replace it.

 

[cliffordcooley]

There is no need to replace it.

The idea was dependent. That is not suggesting replace. Windows doesn't support many other formats, so therefor it is dependent on only those formats it does support.

 

[Puiu]

NTFS is one of the most stable and well known journaling file systems in the world. There is no need to replace it.

It's stable but definitely not the most stable. NTFS was saved by SSDs which help alleviate some of the big issues like fragmentation.

the closed nature of ntfs also means that it takes ages for new modern features to be added and it also has to be implemented with hacks and emulation in other systems (a big reason why it is slower on linux)

the biggest reason why ntfs is still a thing in 2019 is because windows. was built around it (for example the registry is dependent on NTFS transactions).

At least they added DAX storage support in windows 10.

 

[retsxel]

The idea was dependent. That is not suggesting replace. Windows doesn't support many other formats, so therefor it is dependent on only those formats it does support.

However it wouldn't take much to make Windows compatible with an alternative.

It's stable but definitely not the most stable.

In what way?

 

[cliffordcooley]

However it wouldn't take much to make Windows compatible with an alternative.

I think that is what Puiu is saying. But for some reason Microsoft is stubborn on many fronts.

 

[FF222]

It's stable but definitely not the most stable. NTFS was saved by SSDs which help alleviate some of the big issues like fragmentation.

For one, stability has nothing to do with fragmentation. And the second thing is: fragmentation was actually one of the plus points of NTFS compared to competing file systems. Also fragmentation is unavoidable under all file systems.

Some file system drivers do background defragmentation, but that doesn't mean that those file systems don't get fragmented - actually, if you think about it, it means the very opposite, ie. that they get very fragmented, and hence the need for continuous defragmentation in the background. And Windows now also does automatic and periodic defragmentation of NTFS volumes.

the closed nature of ntfs also means that it takes ages for new modern features to be added

Again, this is a non-sequitur. The source being closed has nothing to do with the speed at features are added. And just because something is open source doesn't mean that it magically develops itself at an insane speed. Instead the speed at which something evolves is determined by how expectations evolve and how much manpower is actually thrown at developing it. Regardless of whether its source is open or closed.

and it also has to be implemented with hacks and emulation in other systems (a big reason why it is slower on linux)

It's slower on Linux because there's less manpower thrown at the Linux implementation of the file system driver, not because of any particular property of NTFS.

the biggest reason why ntfs is still a thing in 2019 is because windows. was built around it (for example the registry is dependent on NTFS transactions).

No, it's not. It's dependent on the file system driver providing atomicity. But it does not need the file system to be NTFS, because it uses no feature that's somehow specific to NTFS.

At least they added DAX storage support in windows 10.

Like PM wouldn't be a totally new thing to PCs.

 

[retsxel]

I think that is what Puiu is saying. But for some reason Microsoft is stubborn on many fronts.

In the case of NTFS, I have to agree with them. There is no need to fix/replace what isn't broken, works perfectly and will continue to do so for the foreseeable future.

 

[Igrecman]

I love ext4
It never needs to be defragmented.
Is it because of Linux or because of how partition format works that it never fragments?
And why can't Windows use that format as well? Thanks.

 

[FF222]

Ext4 needs defragmentation just as much any other file system. File system fragmentation is virtually unavoidable on all file systems and in all cases where non-predictable amount of data is written to the disk in a non-predictable pattern - especially if done so in a parallel execution environment. Which pretty much sums up every generic purpose computer system in existence.

Windows would not gain anything by using ext4, and it would actually loose some of features that NTFS has but ext4 doesn't. It's not like ext4 > NTFS or something. That said there's actually a 3rd party ext file system driver for Windows, which can also access ext4 file system. It just has no use or practical advantage for 99% of Windows users.

 

[retsxel]

I love ext4
It never needs to be defragmented.

That's not exactly true...

Is it because of Linux or because of how partition format works that it never fragments

Kinda half way inbetween. Linux is very efficient, yet it does have fragmentation issues. Linux works in the background to manage fragmentation, so you never actually notice it. If a swap partition exists on the system this is great reduced. That applies to mechanical HDD's however. SSD's are handled very differently and fragmentation isn't a problem where wear-leveling is. But I digress..

And why can't Windows use that format as well? Thanks.

That one is simple, Microsoft wants to keep some control.