Think the Windows operating system is expensive? Wait till you heap on the additional costs for the third party software required to secure it, including virus scanners, spyware checkers and firewall managers.
If you still have something left in your budget, you may want to shell out for registry cleaners… or not, if you can get them for free.
Just to prove it can be done, Email Battles set out on a shopping spree for top-notch registry cleaners. Maximum allowable price: Zero. No demos or shareware.
As many network managers must feed the whole family of Windows systems, we insisted that the registry cleaners must run on Windows 98, ME, XP, NT and 2000.
Two products, CCleaner and RegSeeker, met our criteria. And each brings something special to the table.
CCleaner is better known and nicely polished. It deletes unused or obsolete application paths, applications, ActiveX controls, classes, desktop shortcuts, file extensions, fonts, help files, installers, missing shared dlls, MUI cache, software, and Start Menu shortcuts.
It’s a no-brainer to use. (Has a lot of other features, too, like cleaning out your Internet cache and cookies.) And it’s really fast.
CCleaner’s registry checking speed is achieved by foregoing a thorough directory scan. That’s where RegSeeker comes in. It takes much longer because RegSeeker scans both the directory and registry, then compares them. That’s why it digs up more trash.
We ran both products on a shopworn Windows XP and a beat-up Millennium desktop. On the XP, we ran RegSeeker first, while CCleaner took the opening slot on the ME. Results were similar on each system. RegSeeker cleaned around 900 entries. CCleaner netted 30 to 40. Each product picked up trash the other missed. (fyi: A well-known commercial product, Registry Mechanic, found just 300 bogies.)
In both cases, we blindly deleted everything the scrubbers recommended.
Results: On reboot, everything worked, and both machines ran markedly faster.
CCleaner is unambiguously freeware. On the other hand, RegSeeker’s licensing is a bit tougher to figure out. While the vendor says RegSeeker is free for personal use only, the banner on Hoover Desk’s online store says “We regret to let you know that our site is currently closed.”
We’ve dropped a line to the vendor. We’ll let you know what we find out.
Until then, we’re going to continue playing with both.
Oh. Once you’re done scrubbing a registry, you’ll want to compact it. Try WinASO RegDefrag. It physically defragments the Windows registry file to further speed up your system. Price is right, too, i.e. free.
Have a favorite reg cleaner? Let us know what you like and why.

10 comments
Comments feed for this article
January 31st, 2006 at 11:13 am
APK
I would like to compare a registry cleaner I wrote years ago vs. the one you used & the others you tested against as well!
(Mine operates on the same principle that your article describes RegSeeker operating on & mine has been in existence since 1998 or so/iirc)
I would like to submit it for testing comparisons.
BUT, first, I would like to know the mechanics of the test methods used first however.
I.E. - do you or any of the programs in the suite first “inject” a .reg file into user’s systems prior to testing?
If that is your test method, I have to disagree with it, & do not wish to participate.
(This is simply because I am aware that since diff. programs of this type in their engine/algorithm CAN & DO favor certain data types ‘injected’ into the tested system, such rigging of a test is not acceptable as a fair one).
E.G.-> YEARS ago, circa 2001-2002 iirc, I saw the same thing years ago from JV RegCleaner as to testing methodologies!
(The programmer of that software, a Juoni Vuorio, was doing that very thing (inserting a ‘jury-rigging’ test .reg file into user’s systems FIRST, & then running his program against all other comers, rigging the test via a preset data set that his program obviously favored)).
Later, I had users run HIS program vs. my own (”APK Registry Cleaning Engine”) with their systems & registry data UNALTERED by any such .reg file jury-rigging dataset and ‘lo & behold’, my program outperformed his by 2-3 fold.
I would let the users run BOTH programs (but not deleting what they found, only showing what they found that was legal & safe for deletion, then comparing how many entries these types of programs found as legit to delete.
(I know my program is absolutely safe against damaging a system, because I take out via filters in my engine for my program an area I found which was dangerous to give users the opportunity to delete, & that is OLE Server registrations - burn the WRONG ones? Things stop working for example).
Above all, Remember - I had users do their OWN tests on their own systems without altering their data in the registry first via any .reg file injection of a ‘test set’ (jury-rig results set is more like it).
So, I have to ask on your testing methods used first.
If you are NOT doing such a thing?
Then I would gladly submit my program via email contact on your parts to myself, so I can send you my program to compare it vs. the ones you have tested.
(APK Registry Cleaning Engine has a VERY simple installation you can perform manually with easy directions, just single .exe & 1 .dll you copy to %WinDir% & %WinDir%\system32, no OLEServer registration required etc. & run it from a shortcut to the .exe file).
Personally, I would like to see how my program performs against the competitors out there today/nowadays, provided the test is fair per the constraints I laid out above.
Thank you,
Sincerely,
APK
apk4776239@hotmail.com
January 31st, 2006 at 11:59 am
APK
For the tester’s reference, a test performed by this gentleman which you can inquire with him about (he has given me written permission to allow this & has done so in the past regarding others testing my registry cleaner vs. those of others out there):
Phil Atherton [tecknowledgy@btinternet.com]
Whom you can write via email & inquire with regarding this testimony of “APK Registry Cleaning Engine 2002++ SR-7″ (latest model) vs. others out there & his test methods + findings, quoted below from March 14th, 2005:
——————————–
Alexander:
I’ve built 2 machines, WinXP Pro SP1 on an Athlon 2800+ system and MediaCentre 2005 on a 3.0GHz P4 system. I had installed Regcleaner 4.3 by Juoni Vuorio on both machines.
Regcleaner didn’t work on my MediaCentre PC (it is an OEM machine which I have rebuilt a number of times and tried Regcleaner each time but I suspect there’s some OEM gubbins in the Windows set-up that’s causing some instability) but your cleaner worked on both. I’m not surprised with the APK simple install arrangement.
I regularly clean my machines using simple tools, Regcleaner, Norton System Works, X-Setup Pro, Windows Defrag, etc.
After running these I immediately tried the APK cleaner to see whether it picked up any additional registry errors, and it found 39! It regularly finds approx 20 more that Regcleaner and takes approx 18 minutes to do the job to Regcleaner’s 5 mins. This is what I would expect as I have a large number of apps installed (approx 36 including Office XP Pro, AutoCAD 2004, etc) and is not surprising if the cleaner is to do its job properly.
I do believe I can tell the difference in the operating speed of the machine afterwards. (Slight)
I am impressed.
Phil Atherton
Senior Consultant - Communication Systems
——————————–
If you have ANY questions regarding my program, write me by email (address submitted to post here, or just use the one @ the bottom of my posting here) & I can send the program (726kb zipfile with manual install/uninstall directions + .bat file automated installers which you can edit & understand easily to amend them to your own unique install areas (the .dll MUST be a publicly visible one in %WinDir%\system32, & the 2 .reg files for backup tracking of data deleted from the registry must be in %WinDir% & %WinDir%\system32, but the .exe can run from anyplace on your system & is fully self-guiding/intelligent in its operations).
BUT, should you have ANY questions? Feel free to write me @:
apk4776239@hotmail.com
APK
P.S.=> I am truly interested in seeing how my work compares with those today, featureset-wise + safety + ease-of-use-wise, as well as accuracy (provided, as I stated above, none of the testing involves injecting .reg files into user’s systems first, rigging the results in such tests - this IS possible & I proved it years ago in head-to-head tests vs. another program of this kind, JVRegcleaner)… thank-you! apk
January 31st, 2006 at 6:15 pm
Editor
We used two in-service computers that are regularly put through all manner of angst with very little mtce.
You can compare your product with those in the article by simply following the links and downloading the software.
Good luck.
February 1st, 2006 at 7:50 am
Rakhesh
I checked up WinASO’s RegDefrag, its not free. All they have is a 30-day trial program, and that too has limitations. http://www.winaso.com/index.html
February 1st, 2006 at 4:34 pm
Editor
From the publisher’s site:
— Full Free (FreeShare) —
WinASO RegDefrag Version 1.2
Update: Dec 05, 2005 Version: 1.2.
(www.winaso.com)
February 1st, 2006 at 4:41 pm
je.saist
I used to be in favor of jv16’s powertools program, until it went proprietary
thanks for these listings, now I have something that actually appears to work
February 25th, 2006 at 1:42 am
bill harrison
just wanted to say thanks for the regseeker link, there were allmost 1900 issues i deleted on my old windows 98se pc. i,m a die hard and was ready to scrape my old p3 but maybe now i,ll get another year or so out of it, i,m definitly running faster , thanks
August 18th, 2006 at 12:23 am
pancho
I know that one of the criteria for this article was that there was a budget of zero for the applications that would be considered and I can respect that. I do however have a opinion about what you get for free compared to what you pay for. I personally have purchased and use StompSoft’s Registry Repair and I feel that no freeware application can truly stand up to a fully developed commercial version. At least that’s what I think.
August 18th, 2006 at 7:25 am
turbo37
pancho:
Yeah, people who run their entire businesses on free software like Apache, MySQL, Linux, FreeBSD, PHP, etc. sure are idiots!
Uh… NOT!
September 7th, 2006 at 12:14 am
pancho
turbo37:
Alright you got me on that.
Though I’m sure you would have to admit that there are times when purchasing the right software is good for business as well. (Because not everything is life is free)