Hacker News new | past | comments | ask | show | jobs | submit login
Ongoing MacKeeper fraud (thesafemac.com)
77 points by zdw on Nov 2, 2014 | hide | past | web | favorite | 28 comments



As a Mac user (and someone who spends most of the day in Xcode), I feel lucky to have a machine that generally works most of the time and I don't need to defrag it, virus-scan it, memory-clean it, watch running processes for free RAM, and so on and so forth. So, I feel really bad for users who feel duped into downloading things like this -- they get a subpar computer as a result, and they totally don't have to.


The #3 app in the mac app store is "Memory Clean", and apparently it was featured by Apple as an "invaluable utility" and won a Macworld award.

This is really strange to me because, like you, I thought OSX's native memory management ought to be optimal and tools like this are just placebos.


I've tried to observe how this app works.

It actually forces OS X's native memory management to kick in.

It does this by using an obscene amount of memory for a very short amount of time. This forces OS X to clean up the existing RAM in use as the RAM used by Memory Cleaner builds up dramatically.

In this way its partially cosmetic since your RAM would be cleaned up if you used more of it anyway.


There are tricks that iOS Devs (including me) use to make iOS cleanup after itself, for example iOS can take its sweet time releasing memory from MKMapViews even after the controller is gone so a trick is to switch the Map Type just before the controller is released as it forces iOS to drop the cache for the current map type. So yeah, I've not used that program but I assume it's using similar tricks in OSX.


But that's talking about memory usage tricks within a process, which is a totally different beast from what this app claims to do.


So what exactly is the problem with it taking its "sweet time"? Isn't that just a way of saying it will wait until it actually needs the memory to drop the cache?

Perhaps if there was a significant performance hit incurred when dropping caches you may want to do it more eagerly.. but is there?


Memory management takes the form of optimism, they'll not clear down memory in case the user asks for the same thing again. For people with lots of RAM (8GB+) this is fine.

However if you have a 4GB computer and generally single task, then a memory clean tool can work in your favour.

I have several virtual machines that start up overnight and close down in the morning, at that point I want my memory flushed to start a fresh with my applications not the VM's cached data, which wont be needed again for another 12 hours.

This is of course a corner case and just last week I upgraded to 16GB of RAM mitigating the requirement. But there is no one size fits all memory manager out there.


I use "sudo purge" from time to time. I wonder how this app is different/better.


purge(8) flushes and clears the disk buffers, which is not the issue that kills me with my old macbook with 4G ram.

It looks like disk cache expiration is slow and tends to kick in when it's a bit too late. With the speed that modern apps chew through memory, I'm sent into a swap thrash and get sluggish response and beachballs.

MemoryKeeper on the App store seems to mostly clear the file cache and do some -fu to compact memory (maybe it's a memory defragment, the docs are a little lacking in technical details.) When I manually run it, everything runs great until the free memory dries up and I'm back in that alloc/swapthrash/cachepurge cycle again.

I wrote a tiny C program for linux to force cache expirations (in a very sloppy but reasonably portable way) back in '98 that I've started using again on osX, it fixes the issue and runs quite a bit quicker than MemoryKeeper. I can make it available if anyone else wants it.


I'd like to see it. (my email's in my profile.)


Never run a virus checker (there are none) and I don't download stuff I don't know (to avoid malware of some kind). I have no clue why people think they need something, especially one with such a scummy background. I wish Apple would simply block the app from running to save those who fall victim.


Well it doesn't appear to be on the App store, so I suspect that most inexperienced users won't be able to install/run it in the first place, due to OSX's default security permissions.


I run Claxmav[1] on all my macs because one way or another malware will reach my computer, so I'd better be prepared. I never had problems since 2004, never had a mac virus or anything but that doesn't mean they don't exist.

I think to be sure you need a combination like: ClaXmav (monitors predefined dirs and incoming mails), LittleSnitch (monitor outgoing connections) and SpamSieve to manage spam with Mail.app.

[1] http://www.clamxav.com


How exactly do you think they get duped into downloading this in the first place?

It's certainly not because they "feel lucky to have a machine that generally works most of the time and I don't need to defrag it, virus-scan it, memory-clean it, watch running processes for free RAM, and so on and so forth".


Scare tactics, the same way unethical car mechanics sell useless tune ups and repairs to the clueless. Or the Indian call centers claiming to be ms/dell/hp support and that the callee as a virus they'll remove for 99$. Some of my elderly relatives actually fell for that and then shame delayed getting help.

It's despicable.


It's FUD. They don't see anything wrong, but they don't know much about computers. Something could be wrong, so better download this too just to be sure.


Quite easily. I almost did quite recently as 'mackeeper' was offered with sf.net's 'installer' for Filezilla.

It's almost impressive, in a sad sad way. They've pretty much cornered the market on mac malware. Unfortunately they are the malware, not the solution.


You can tell a lot about the legitimacy of a company in the "security space" by their SSL/TLS config on key sites. As of this reply they've still not patched / disabled SSLv3 to guard against POODLE.

https://www.ssllabs.com/ssltest/analyze.html?d=store.mackeep...


You should not blindly believe ssllabs.com without spending time further digging into the results. For example, some sites have decided to keep sslv3 but only support ciphers which are not prone to Poodle.

I'm not saying it's the case here, but be cautious about blindly believing everything ssllabs.com reports.


Where sites have chosen to keep SSLv3 and mitigate via cipher selection SSLLabs makes note of it without impacting the overall score: "This server uses SSL 3, with POODLE mitigated. Still, it's recommended that this protocol is disabled."


I was going to say it's not always the case, but you are right! Thanks for pointing that out.


I'm 100% sure that in the past MacKeeper was recommended on support.apple.com (not discussions.apple.com).

As far as I remember it was a page with three suggestions, not unlike [1] and the last one was to install MacKeeper. MacKeeper was just mentioned, but not linked which appeared odd to me at the time. This was about two years ago.

Unfortunately I couldn't find the page anywhere, not even at archive.org.

[1] http://support.apple.com/en-us/ht1147


I always get mackeeper popups when I'm on porn sites.


Yes like everybody else. A friend of mine, asked me twice, if this MacKeeper will keep his mac safe. I told him it's a simple spyware and he has to avoid it like the plague.

It's kinda widespread software among mac users, because you can find banners literally in every website with disputable content like porn, torrents, subtitles, you-name-it.


I wonder if anything like that, but targeted at linux, will ever come up.


It would have already if there was the market share.


Hm hardly. The average linux user is something like 100 times more skilled, computer-wise, than your average windows/mac user. That's why IMHO it's highly unlikely that a Linux Desktop based malware will ever reach big numbers. Those linux people are mostly geeks, they can tell easily if something is going wrong...


I disagree.. If linux had a larger market share you'd see users that weren't '100x more skilled' and advertising a cleanup program would be viable.

Permission structure of *nix is the same, and fools can be fools anywhere. A 30% market share for linux doesn't mean 30% of computer users gain 100x the skill overnight.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: