Comparing Registry Cleaners

[arraknid]

There's been some recent discussion on the forum about Registry cleaners and it seems that many people consider them to be some kind of cure-all for slow machines or intermittent problems. Looking at the availability and marketing of such programs, it's easy to see why. A quick search found nearly 70 different offerings, including 25 or so free versions, though some of those offer free scans only. Curing the so-called 'problems' then costs money.

There are many sites offering comparisons of Registry cleaners - though most, it has to be said, are far from independent and their conclusions should be viewed with some degree of suspicion.

So, are they actually necessary? Personally, I'm with those that have their doubts. This is a quote from Wikipedia...

Most notably, critics say there is no reliable way for a third party program to know whether any particular key is invalid, redundant or neither. Poorly designed registry cleaners may not know for sure whether a key is still being used by Windows or what detrimental effects removing it may have. This has led to examples of registry cleaners causing loss of functionality and/or system instability.

That said, problems can occur when Registry entries become corrupted or changed by malware or viruses, though I've yet to see any product that will cure those problems. Most products use a set of algorithms - effectively a set of instructions based on rules determined by the writer, and as every product is different, so will be the results.

There seems to be a view that the more 'problems' a program finds, the better it is. Again, that would depend on what the software writers have set as the search criteria. Some programs are much more aggressive than others, though whether they need to be is questionable. I get the feeling sometimes that the marketing of the product demands such fantastic results to supposedly raise it above others in the market.

It is true that uninstalled programs have a habit of leaving redundant entries behind which will eventually clutter the Registry, and cleaning it from time to time can only help. Whether it speeds up the operating system is debatable. It's also true that, as with a hard drive, data becomes fragmented - hence the appearance of products that will defrag the Registry. Again, in theory, that should speed things up, but probably only by a few milliseconds.

The majority of Registry defrag products run within the Windows environment which is hardly ideal when there are so many programs running and needing to access the Registry. The ideal option is to defrag before Windows loads, in the same way that chkdsk does. My preferred option is to use PageDefrag which does just that - either at each boot up, or on a one-off basis.

I recently had to clean out a machine loaded with XP SP2 after they clicked on a nasty. With EF's help, it's now clean, but as it has never had even a defrag in 4 years, I thought I'd run a few tests comparing some Registry cleaners. The choice is based on what I had available so isn't in any way definitive, merely something to satisfy my curiosity.

I only compared four products, because to have an accurate comparison, I needed to restore each system from a backup image after each test, and that takes more time than I had available.

Products tested were Registry Mechanic, Ccleaner, Eusing and RegCure - each one in its latest version. The first stage was to simply run the scan options to see what the results were. The number of 'problems' found was as follows....

Registry Mechanic 1678
Ccleaner 963
Eusing 658
RegCure 858

Next stage was to run the fix option and restart. Results as follows....

Registry Mechanic
Refused to start reporting config/system error

Ccleaner
Started up OK, most programs ran fine, but froze when IE was started.

Eusing
Started up and ran OK. Tried most programs without problem.

RegCure
Started up and ran OK. Again no problems running programs.

Trying to compare what each program had actually done would also have taken far too much of my time, but the wide variation in 'problems' found would tend to confirm a more aggressive set of algorithms in Registry Mechanic and Ccleaner, leading to subsequent system problems.

Comparing the last 2, Eusing and RegCure, they performed much the same, being less aggressive, but one is free and the other isn't! That's not to say that under different circumstances on another machine, the results would be the same. It would depend on the system configuration and what software was installed.

Hopefully, what the tests achieved lends weight to the argument that Registry cleaning software can cause more problems than it cures, so should always be used with caution.

Let me add that I have no interest, commercially or otherwise, in any of the products tested, nor should the test results be seen as favouring one product over another.

I'm simply reporting what I found.

[netserf]

Using CCleaner frequently the past year I've never experienced a problem. It is simple, and easy to use. I have no connection with the people or company providing the software, and no financial interest in posting this information.

[Canuck]

Are we talking Ccleaner's regular cleanup utility or Ccleaner's registry cleaner .. two different animals, but in the same program.

[MrDarn]

I've had mixed results with Ccleaners registry cleaner, and now no longer use it.

on my desktop and laptop, no visible speed was gained, but everything worked, however i've now used it twice on a packard bell laptop, with disastrous results, namely the laptop not booting to windows after the restart.

[redoak]

These comments come from another "help" site.

{redoak}

Posted by "Elvandil," a Moderator with over 30,000 posts:
Never use any of them. They are scams, do nothing but slow down the machine, and remove functionality that you may not miss until later when you need to install new hardware or repair the system.

The "errors" they find are not errors at all, and over 90% of the "fixes" they make are incorrect, but the programs then see those new errors as "correct".

They are based on a false assumption about how the registry is accessed, and the also false idea that all registry entries should follow patterns and any that don't must be "errors".
- - - - -
Advice and Info re Registry 'cleaners'- excerpts from TSG, 10-07

The main reason they (are apt to cause problems) is because the cleaners do NOT just clean up stuff left over from uninstalls. They also "guess" that other items need to be deleted, many times incorrectly. They also attempt to "fix" references to files that no longer exist.

As a result, you can end up with programs that no longer work, functions that no longer work and in some cases a PC that will no longer boot.

There is no performance gains to be had by running these things. Windows does not read the entire registry when it is looking for something. It uses the keys which allow them to go straight to the item that they are looking for.

References:
Why I don’t use registry cleaners | Ed Bott’s Windows Expertise |
Windows Registry help


If you really want to see how active the registry is in your normal running environment, download and run Regmon.

http://www.microsoft.com/technet/sys...on/Regmon.mspx

As you will see, the registry is constantly being accessed. If, for each of these, the entire file was being searched, your PC would be amazingly slow and not responsive.

[GavnBridg8]

There is an article on a Macecraft Software site that reports on their comparison / evaluation of 31 registry cleaners. The list includes free and paid-for cleaners with many recognisable names in the assessment. Hope it is helpful.

I'll leave you to see which was the 'winner' but it raised a question of objectivity for me.

Link from reimage.com to article -

Study: Registry Cleaners Cause more Harm than Good

Macecraft article URL

http://www.macecraft.com/registry_cleaner_comparison2

[Canuck]

Hi GavnBridg8, Thanks for your input, however if I'm reading this right, the report from reimage is a quote from Macecraft who produce JV16 which is a registry cleaner. I would question their objectivity in this study and much prefer to rely on studies done by those who don't have a vested interest in the products they review.

[evilfantasy]

Reimage has some mixed reviews. reimage.com | WOT Security Scorecard | WOT Web of Trust

[arraknid]

This is yet another 'independent' study of Reg Cleaners, which is so biased towards their own product, it makes a nonsense of the entire study.

The best advice is NOT to run ANY registry cleaner, any time, anywhere!!

[Osc]

This is a great topic - thank you for bringing this up arraknid. I wrote a little article for this to sum up what was said here and add my own thoughts.
Help2Go - free computer problems help and troubleshooting - Registry Cleaner Review
If I've missed anything, please let me know so I can amend the tutorial.