CleanMem 2.4.2

Changes in CleanMem 2.4.2:
- Cleaning of the File Cache is now disabled by default. There are times where cleaning it helps improve performance and sometimes it doesn’t depending what the system is doing at the time. So best to let the user decide to have it enabled.
- Added a new menu option to the mini monitor to manually clean the file cache. This new option is part of the free and pro versions
- CleanMem and all of its exe’s are now terminal services aware.

http://www.pcwintech.com/cleanmem

http://www.pcwintech.com/about-cleanmem

 

Wait wait wait. What exactly does this program do? It looks to me like a RAM cleaner... i.e. snake oil.

Modern operating systems use RAM not occupied by programs as filesystem cache, which you don't want to clear out because it makes it faster to access files. On Windows NT (or OSX, or Linux) you want to keep as much stuff in RAM as possible. The more filesystem cache you can have at any time, the better.

There are situations (on Linux anyway) where overly aggressive filesystem caching causes running programs to get swapped out, but you wouldn't solve that by clearing the filesystem cache every 15 minutes; you'd solve that by telling the OS to reclaim filesystem cache faster. Purging the filesystem cache is just going to make your computer crawl.

 

Read on: http://www.pcwintech.com/about-cleanmem

 

It gets lumped in with all the snake oil memory 'optimisers', but it works quite differently.

On a system with plenty of memory, you won't notice a difference one way or another - unless there's a memory leak in a program, in which case it can be demonstrably useful.

On an old system with low memory, it's the next best thing to upgrading RAM and rationalising autoruns. Subjective experience is less page file usage/disk churning.

It's just one tool in the kit.

Some people are 100% against all memory optimisers, but it's not something I can be bothered debating about. Even detractors admit that CleanMem doesn't affect negatively performance.

 

I see. If the description is accurate, this is the Windows equivalent of e.g. setting swappiness to something low on Linux.

Aren't there registry settings to control such things though? I have trouble believing Windows doesn't some built-in method for tuning swappiness, or rate of filesystem cache reclaim. Running a scheduled task or service to deal with swapping problems seems like huge overkill.

Also, in my (limited) experience Windows is actually very good about swapping. I've never seen a Windows system swap out active programs until almost all available memory is used up. Compare to Linux, which with default settings will swap like crazy when half of available RAM is used by programs.

Finally, I still don't see how this helps with memory leaks. If e.g. Firefox allocates memory and never frees it, I'm fairly sure that memory will not be freed until Firefox is killed. No?

 

Of course it's snake oil.

I've already explained in detail in this topic why such programs (this one included) are either useless or downright harmful.
I asked for some proof that it actually improves performance. I mean, if it really does it should be provable with some kind of tests/benchmarks. All I got was hostility and theories.

 

Oops, wrong topic. I meant this one.
If you still consider that opinion but you consider "it just works" empirical evidence, then you really must live in Oppositeland.

And no, I can't just accept that these programs work. It's like accepting that registry cleaners work or that defragmenting a SSD/flash drive makes sense, or that vinyl is a better medium than CD, or that banks are honest institutions that care about people, etc. Those should be enough examples for you to get the picture.

 

This isn't about opinions, it's about facts. Science and technology aren't opinion based. One can't have an opinion on whether 3 is a higher number than 2.

Whatever, arguing with people who believe such programs are useful is like trying to inject logic and reason into religion.

 

Night Raven, your "arguments" were destroyed here. You didn't reply to those points.