View Single Post
Old 01. May 2009, 09:54 AM   #5 (permalink)
chris.p
Foundation Editor
 
chris.p's Avatar
 
Join Date: Apr 2008
Location: Kent, UK
Posts: 1,634
Default

Well, I suppose I better give a sensible answer.

I cannot see that compressing the registry will make any difference unless you have a 500GB disk, hundreds of apps and millions of files. If that's the case then OK, you might see some sort of improvement. But I reckon you'd get a better speed increase by managing your disks and data better, if you have that much.

I have used several registry cleaners and found absolutely safe ones (in the right hands), through to lethal ones. The only one I trust now is a really old version of Ontrack Fixit Utilities, that I've used hundreds of times on different PCs without a single issue ever. But it's unusual: it classifies dead keys as green, yellow or red, and although you can delete the green ones, the others you must check each one manually. You must also rebuild all the links manually, for the faulty ones - it doesn't just delete them like all other apps do as far as I know. I've never seen this on any other app and I assume it's why it can delete and rebuild 2,000 faulty reg keys on a well-used machine without a single issue. But the problem is the time - you manually fix the dead links. It asks you every time what the link target should be.

However the result on the worst machines like heavily crudded-up ones with 2,000 faulty keys or so is a noticeable speed increase in booting up and in multi-tasking jobs. I've noticed this many times and it's not a fantasy. But with 400 dead keys I wouldn't expect to see any measurable increase.

All other apps I've used resulted in some sort of issue, perhaps because they are so fast, and ask for no logic input, and because I went too far with them. So I know at least one works, and works safely, on heavily used machines where a lot of apps are installed and uninstalled, and where the cleanup app works very slowly and with an extended manual input.

That's my experience. Now, I know of a researcher in this field who emailed more than 30 registry cleaner software authors and specifically asked them for some sort of proof that registry cleaner apps work. In other words - what precisely was the benefit and how was it measured. All replied with various excuses or other platitudes, and no explanations. He emailed again, explaining that if they couldn't reply with any kind of evidence, there was a risk that his review might be biased. Again they could not provide any tests or measurements. So as far as the industry itself is concerned - or so I'm told - there are no test results available from any providers that show any case for the use of registry cleaners.

However my own experience, as stated, is sufficient for me to use one *in certain circumstances*. Otherwise - your time and trouble are probabably better off spent elsewhere.
chris.p is offline   Reply With Quote