[rms]

Jeremy, on Nov 11 2006, 09:12 PM, said:

You don't need evidence to know that the need to defragment drives is greater than ever before, you only need common sense. More and more people are downloading archives, movies, albums. More people are using video editing software, trying to bring massive systems to their knees, more and more Read/Writes! With more of this happening, defragmenting needs to be done more often (constantly) to keep drives at their peek peformance. Diskeeper 2007 does this without needing you to do anything.

Common sense is of course a good thing, if it is not affected by defrag software vendors advertising.
What about disk drives ever advancing caching algorithms and growing cache size, which reduce some effects caused by fragmentation? What about NCQ, which addresses other effects of fragmentation and does it quite effectively?

Of course, it does not eliminate the need to defragment drives, but to say that this need is greater than ever before, is a bit of loud statement, which may need some clarification, or, indeed, evidence.

And, what about improvements in Windows system tools, like - built-in defragmenter now optimizes layout.ini files every time manual defragmentaion pass is run? Quite funnily, this fact is "unknown" to experts like Raxco technical suport, and Diskeeper is not advertising it either.

On the other hand, common sense should tell us that there are no "invisible" services, even if they may be named named like "InvisiTasking".
It may be difficult to demonstrate the adverse effect of those "invisible" things on the normal home computer (the same way it is nearly impossible to prove the advantage of third-party defrag tools over built-in defragmenter on such computer - except that lazy user doesn't need to press the button once a week), but it is quite easy to show it on the audio workstation. There you can reduce latency from about 256 samples to something like 64 samples, simply by turning off those "invisible " things.

Finally, but this of course applies only if we are really considering paying $100 for defrag software - for that money one can seriously upgrade system and enjoy real perfomance gain, not one based mostly on placebo effect.

[jcarle]

rms, on Nov 15 2006, 10:30 PM, said:

Common sense is of course a good thing, if it is not affected by defrag software vendors advertising.

Advertising has nothing to do with it considering this thread is about user experiences with their own favorite defragmentation utilities. I see no marketing tools being used anywhere in this thread.

rms, on Nov 15 2006, 10:30 PM, said:

What about disk drives ever advancing caching algorithms and growing cache size, which reduce some effects caused by fragmentation?

Reduces but does not eliminate.

rms, on Nov 15 2006, 10:30 PM, said:

What about NCQ, which addresses other effects of fragmentation and does it quite effectively?

Native Command Queueing has nothing to do with disk fragmentation, does nothing to help prevent it and does nothing to reduce the effects of disk fragmentation.

rms, on Nov 15 2006, 10:30 PM, said:

Of course, it does not eliminate the need to defragment drives, but to say that this need is greater than ever before, is a bit of loud statement, which may need some clarification, or, indeed, evidence.

The evidence has already been brought fourth, a simple count of the number of files on a typical modern computer is more then enough evidence in itself to justify defragmentation. Considering that 8GB hard drive were landmarks less then 6 or 7 years ago, 320+GB hard drives filled to the brim with thousands of files are in much greater need of defragmentation then a mear couple hundred files that used to fill computers.

rms, on Nov 15 2006, 10:30 PM, said:

And, what about improvements in Windows system tools, like - built-in defragmenter now optimizes layout.ini files every time manual defragmentaion pass is run? Quite funnily, this fact is "unknown" to experts like Raxco technical suport, and Diskeeper is not advertising it either.

Well, it's unknown to them surely because it has nothing to do with disk defragmentation. The prefetcher has no corolation to disk fragmentation. And in fact, I quote from microsoft's own text regarding the prefetcher:

Quote

Disk Efficiency Optimizations
The physical placement, or layout, of files on the disk can have a considerable effect on performance, up to 10% for normal use. Windows XP observes file usage patterns as the system is used. If deemed necessary, Windows XP will adjust the file layout at three day intervals. By placing files that are referenced together near each other on the disk, and towards the more dense outer edge of the disk, seek distances are reduced which results in shorter seek times and improved performance. The performance benefit of placing files becomes increasing important as the size of the disk increases.

The files moved for more efficient layout are also kept contiguous. Windows XP does not intentionally fragment files as was done by some earlier versions of Windows.

Even though the disk layout optimization does insure some files will be defragmented, it is not a complete substitute for fully defragmenting the disk. Users should still defragment their drives regularly. The built-in Windows XP defragmentation program understands the file layout directives and will position the files if they arent already placed properly. However, it will not update the layout file with new information gathered in the last three days. Without manual intervention, the layout file, %windir%\Prefetch\Layout.ini, will only be updated once every three days. The contents of the Layout.ini file can be viewed with Notepad.

Source: http://www.microsoft.../benchmark.mspx

Do take special note of the phrase "Even though the disk layout optimization does insure some files will be defragmented, it is not a complete substitute for fully defragmenting the disk. Users should still defragment their drives regularly.".

rms, on Nov 15 2006, 10:30 PM, said:

On the other hand, common sense should tell us that there are no "invisible" services, even if they may be named named like "InvisiTasking".
It may be difficult to demonstrate the adverse effect of those "invisible" things on the normal home computer (the same way it is nearly impossible to prove the advantage of third-party defrag tools over built-in defragmenter on such computer - except that lazy user doesn't need to press the button once a week), but it is quite easy to show it on the audio workstation. There you can reduce latency from about 256 samples to something like 64 samples, simply by turning off those "invisible " things.

InvisiTasking does exactly what it says it does. It makes Diskeeper seem invisible in a perceptual sense. It retains it's work for period where the entire system is idle, and stops immediately whenever the system is no longer idle. Hence, giving the perception of not running.

rms, on Nov 15 2006, 10:30 PM, said:

Finally, but this of course applies only if we are really considering paying $100 for defrag software - for that money one can seriously upgrade system and enjoy real perfomance gain, not one based mostly on placebo effect.

Placebo effect? I believe you are seriously misinformed.

This post has been edited by jcarle: 15 November 2006 - 10:19 PM

[Jeremy]

@RMS, if you have a question about newer harddrives and how fragmentation affects them, I suggest you keep everything in one place by posting your curiosities on the DriverHeaven thread located here.

http://www.driverheaven.net/showthread.php...6565#post986565

And the Diskeeper Blog:
www.diskeeperblog.com

And the PerfectDisk Blog:
http://perfectdiskbl...rfectdisk_blog/

This post has been edited by Jeremy: 15 November 2006 - 11:43 PM

[rms]

jcarle, on Nov 16 2006, 12:18 PM, said:

Native Command Queueing has nothing to do with disk fragmentation, does nothing to help prevent it and does nothing to reduce the effects of disk fragmentation.

Quite a funny statement. Normally, you would need to learn the subject a bit, before making such.
But, what about common sense? It was declared present a while ago.
If file is fragmented, read commands will be issued that lead to random read head movements. If these commands are re-ordered to minimize head movements, how is it "nothing"?
Of course, it will happen only under certain conditions(ie. when there is a queue), but in many cases, especially when several files are accessed simultaneously, it can (and does) significantly reduce the effects of fragmentation.
And, this effect is quite measurable, it is possible to test it.

jcarle, on Nov 16 2006, 12:18 PM, said:

The evidence has already been brought fourth, a simple count of the number of files on a typical modern computer is more then enough evidence in itself to justify defragmentation. Considering that 8GB hard drive were landmarks less then 6 or 7 years ago, 320+GB hard drives filled to the brim with thousands of files are in much greater need of defragmentation then a mear couple hundred files that used to fill computers.

Not true. Simple count of files, or even count of file fragments is not evidence.
Couple hundred files is simply not true, and on those 8GB drives were much more sensitive to even slight fragmentation.
Evidence to justify defragmentation is degraded performance, if you can measure it on the real system. Plus, I didn't say defragmentation is not necessary, so it is difference between built-in defrag and miracle tool.

jcarle, on Nov 16 2006, 12:18 PM, said:

...it is not a complete substitute for fully defragmenting the disk. Users should still defragment their drives regularly.".

Well, that is probably exactly the reason why built-in defragmenter exists.

jcarle, on Nov 16 2006, 12:18 PM, said:

InvisiTasking does exactly what it says it does. It makes Diskeeper seem invisible in a perceptual sense. It retains it's work for period where the entire system is idle, and stops immediately whenever the system is no longer idle. Hence, giving the perception of not running.

Really? Well, it may be in the perceptual sense of somebody who would not benefit from defragmentation anyway.
But, on the audio workstation, it is possible to reproduce the following - Select large number of tracks, and when invisible defragmentation starts, hit play button and see what happens, to learn what is "immediately". Or, same with record.
Would be nice if they did that at their miracle works, but that doesn't bring any money, does it?

Finally, I didn't say those programs are bad. What I am saying is, for home user, they are not worth $100.
Probably, PerfectDisk is worth its $39.99 in many cases.

Jeremy, on Nov 16 2006, 01:43 PM, said:

@RMS, if you have a question about newer harddrives and how fragmentation affects them...

I can measure that, even without asking a question, and have already done it.
But you guys, who are trying to make innocent people believe they need to spend $100 on defrag, should ask some questions, indeed.

[jcarle]

rms, on Nov 16 2006, 06:18 AM, said:

Quite a funny statement. Normally, you would need to learn the subject a bit, before making such.
But, what about common sense? It was declared present a while ago.
If file is fragmented, read commands will be issued that lead to random read head movements. If these commands are re-ordered to minimize head movements, how is it "nothing"?
Of course, it will happen only under certain conditions(ie. when there is a queue), but in many cases, especially when several files are accessed simultaneously, it can (and does) significantly reduce the effects of fragmentation.
And, this effect is quite measurable, it is possible to test it.

Let's see... 2GB file with 150 fragments. No NCQ, 150 read commands. NCQ, 150 read commands. Where the difference? Sure, the NCQ may finish faster because of the ordering of it's fetch commands but that's analogous to comparing a 5,400rpm to a 7,200rpm hard drive. It simply does not address the problem, you're approaching the situation using a band-aid fix.

rms, on Nov 16 2006, 06:18 AM, said:

Not true. Simple count of files, or even count of file fragments is not evidence.
Couple hundred files is simply not true, and on those 8GB drives were much more sensitive to even slight fragmentation.
Evidence to justify defragmentation is degraded performance, if you can measure it on the real system. Plus, I didn't say defragmentation is not necessary, so it is difference between built-in defrag and miracle tool.

Compare the number of writes to hard drive back then to hard drives now, compare the amount of data stored, compare the quantity of changes, compare the quantity of files. It is self explanatory that exponentially speaking, defragmentation is more important now then it was then.

You can compare this to a village vs a city. In the village, sure if the one bridge breaks, it affects the village in much more drastic fashion then if one bridge breaks in a city. However, repaving a broken up road in the village for higher efficiency, smoother driving and faster traffic flow has a much smaller effect then doing the same thing in a city.

And you know, no one's claimed that diskeeper or perfect disk were miracle products. Microsoft cannot do everything perfect and they can't be number one in all aspects of every piece of software that exists in the world. If they were, people wouldn't use Winamp, they'd use Windows Media Player. They wouldn't use Vegas, they'd use Windows Movie Maker. They wouldn't use Photoshop, they'd use Microsoft Picture It!. And for the same reason, they wouldn't use a 3rd party defragmenter, they'd use the Windows Defragmenter. Microsoft's not perfect and some companies make a better product then window's own built-in applications.

rms, on Nov 16 2006, 06:18 AM, said:

Really? Well, it may be in the perceptual sense of somebody who would not benefit from defragmentation anyway.
But, on the audio workstation, it is possible to reproduce the following - Select large number of tracks, and when invisible defragmentation starts, hit play button and see what happens, to learn what is "immediately". Or, same with record.

There is no way to reproduce the exact conditions every single time and as such, no way to put the blame on diskeeper. Who's to say that your drive heads were not in different places? Or that another system process was executing a task? Or that the memory manager had decided to move memory to the page file?

rms, on Nov 16 2006, 06:18 AM, said:

Would be nice if they did that at their miracle works, but that doesn't bring any money, does it?

Finally, I didn't say those programs are bad. What I am saying is, for home user, they are not worth $100.
Probably, PerfectDisk is worth its $39.99 in many cases.

Again, no one quoted miracle workings. And your reasoning for the pricing is flawed. Why do some people use Photoshop when Paint Shop Pro is cheaper? Because photoshop does it better and for some home users, that's what they want.

rms, on Nov 16 2006, 06:18 AM, said:

I can measure that, even without asking a question, and have already done it.

I saw your tests. They're flawed in every possible sense. You proved absolutely nothing. All you proved was that there was almost no changes in all three scenarios. You did not replicate fragmentation in all three situations and you didn't even take any measure as to the level of fragmentation before you even attempted to prove the differences. Your data is useless, the tests a waste and the results a flop.

rms, on Nov 16 2006, 06:18 AM, said:

But you guys, who are trying to make innocent people believe they need to spend $100 on defrag, should ask some questions, indeed.

No one's tried to forcefeed anyone. Everyone's shared they experiences with each product and we've move to analyzing the differences between the two leading products, Diskeeper and Perfect Disk to figure out which one is more efficient and more effective. As the representative of diskeeper so elloquently put it. If you don't like it, don't buy it.

And finally, you keep quoting $100 for diskeeper. Yet, diskeeper 2007 home is listed at $29.95, $10 less then your quoted price for Perfect Disk. I see how you like to keep the comparisons fair.

[Zxian]

I've read a lot of what people have said here, and there's a lot of it. Therefore, I'm gonna keep this short.

I use Diskeeper 2007. I have also tried PerfectDisk8. In my experience, both helped system performance compared to no fragmentation (obviously).

Do I know 100% that one product does a better job in terms of system performance? No.

Do I know that one product does a better job of keeping frag levels in check without my intervention? Yes. Diskeeper does this. PerfectDisk requires me to set up a schedule. DK sets auto-defrag on all drives right from the get-go.

Have I seen any downside of not using PerfectDisk? Not really...


When considering which <insert_thing_here> is "better", you've gotta look at the big picture. Sure, if you wanted do, you could compile everything to assembly, and then try to optimize the assembly structure of your program to every little nitpicky bit, or, you can use a higher level language like .NET, Delphi, Java, etc etc etc. Just like here, you could try to argue about every little thing when it comes to defragmenters, but in the end, which is easier to use? In my opinion, Diskeeper has become truly set-it-and-forget-it (the option is on by default in the installer). Install, and never worry about it ever again.


Another thing to point out - time is money. To me, the time I save not having to worry about defragging my hard drive, or settings up a schedule, or whatever makes up for whatever price difference might come up between PD and DK.

[Jeremy]

Well said, Zxian. Well said, indeed.
RMS, take a look at the thread I linked (now far above) if you have any desire to nitpick any further. If anyone has dragged defrag program debating around, it's me. Zxian used to smack me around about it in the past.
Ease of use and time-is-money are the main points. Just let it go...
Now let's all go play games and tickle our younger siblings if we have any.

[Woomera]

i really dont mean to bother anyone by bringing this topic back up but i couldnt keep myself from saying "THANK YOU AGAIN SO MUCH JEREMY" for the refrrence and infos(though i got nearly a headaches somewhere in middle of page 3 of the topic).

[Jeremy]

You're very welcome.

[Ascii2]

crahak, on Nov 2 2006, 11:58 AM, said:

Bah. Anything but diskeeper

Agreed. Avoid Diskeeper.

[Jeremy]

Ascii2, on Jan 19 2007, 04:40 PM, said:

crahak, on Nov 2 2006, 11:58 AM, said:

Bah. Anything but diskeeper

Agreed. Avoid Diskeeper.

Why? I've been using it for months. My fragments is 0. My fragmentation level is 0%. What's your logic other than simply agreeing with someone says? Why do they dislike it? Do you even know if your reason and their reason are the same? Explain yourself.

[Ascii2]

Jeremy, on Jan 19 2007, 09:22 PM, said:

Ascii2, on Jan 19 2007, 04:40 PM, said:

crahak, on Nov 2 2006, 11:58 AM, said:

Bah. Anything but diskeeper

Agreed. Avoid Diskeeper.

Why? I've been using it for months. My fragments is 0. My fragmentation level is 0%. What's your logic other than simply agreeing with someone says? Why do they dislike it? Do you even know if your reason and their reason are the same? Explain yourself.

I do not know the answer to all you questions.

Something for you to try: Unistall Diskeeper. Then try to use (defragment) the defragmenter that comes with windows. The defragmenter may not fuction correctly (tested only on Windows 2000 Professional with Service Pack 4). Diskeeper seems to cripple the defragmenter that comes with Windows 2000.

[Jeremy]

Ascii2, on Jan 20 2007, 09:12 PM, said:

I do not know the answer to all you questions.

You mean any of my questions.

Ascii2, on Jan 20 2007, 09:12 PM, said:

Something for you to try: Unistall Diskeeper. Then try to use (defragment) the defragmenter that comes with windows. The defragmenter may not fuction correctly (tested only on Windows 2000 Professional with Service Pack 4). Diskeeper seems to cripple the defragmenter that comes with Windows 2000.

I have no intention of trying this because:
A. I do not use Windows 2000.
B. I have removed the many-years-old obsolete defragmenter function with nLite.
C. As I have stated, DK2007 does everything for me automatically and waits for available resources. No need for anything else.

This post has been edited by Jeremy: 20 January 2007 - 07:50 PM

[Subcodec]

This thread is an interesting read. I actually enjoyed the debate on whether defragmentation software is needed as much as it used to be. IMHO, I believe the answer is yes AND no.

I'm sure many users today have more than one partition and/or drives on their computer. One drive/partition hold the OS, and the others hold data. The partition/drive holding the OS should be defragmented quite frequently as their are always many files that are updated, added, or deleted. That is a given.

But...

The other partitions that hold data, may or may not need defragmentation. For example, on one of my drives, I hold just media. The whole drive has nothing but music and video (technically it's a RAID array, but back to the point). As I add to this drive, I adjust tags and information and then I run one pass with something like Diskkeeper. Two or three weeks later, there is still zero to very little defragmentation since all I've done is read from the drive.

As far as my preference, I still choose Diskkeeper. It has features that I personally like and it also helps that I've used it for so long that I'm comfortable with it. Another small reason is because they are the ones that built Windows built in defragmenter. So if Microsoft can trust the software enough to ship it within Windows, it can't be all too bad. (Somewhat of a oxymoron since Windows itself doesn't always work right )

[slimzky]

ok what services does diskeeper 2007 depends on?

[Jeremy]

Subcodec, a drive that is used for only storage/backup needs less defragmentation than an OS drive. Once a new file is written to disk for storage or backup is made, that must be defragmented. OS drives constantly need defragmentation which is where Diskeeper 2007 comes in handy as it is automatic. I've almost forgotten what fragmentation is nowadays; been using DK2007 so long.

ProjecTK, not sure. I have a lot of Services removed with nLite and then most disabled and DK2007 functions normally.

This post has been edited by Jeremy: 24 March 2007 - 09:37 PM