What are your guys’ opinion on McAfee VPN?

He means it gets bundled with things

Just because YOU don’t accept it doesn’t make it fallacious, just like McAfee is horrible, it might have been good 20 years ago, but it’s definitely not now. You’re either not old enough or don’t know enough about IT my guy. Good luck with your virus

If you see a free mcaffee it’s fake. The real mcaffee starts at 40 a month

They didn’t defend McAfee (and neither am I, I wouldn’t go near it), they just pointed out that your argument is fallacious, which it is - just because ‘everyone’ says X, it does not make X true.

And you guys saying it’s good magically makes it good? Lmfao. Why don’t you tell me and everyone else why it’s good then. I can wait :joy:

I suggest you re-read the first half of my comment…

So then what does it take to make something “bad” besides the entire user base saying it’s bad? Like if everyone saying something is something then it actually does make it that… how else do we describe something? McAfee is bad, we know it, they know it. People just have to be “those guys” apparently and make a big tissy. Congrats on make mf the tissy mate :roll_eyes:

Depends what you mean by bad. Generally when looking at AVs there are 4 main aspects looked at:

  1. detection/protection
  2. performance impact
  3. UI/UX
  4. cost

The latter 2 are subjective, or at least relative to other AVs. These are really the only cases in which ‘everyone says so’ would actually be a valid argument, were it actually true.

You could also throw business practices in there as well, which is my main reason for avoiding McAfee (also Norton and Avast/AVG). If you want to read more about why I would never use McAfee or Norton, read here - https://www.reddit.com/r/antivirus/comments/y679ni/comment/isnrm9o

There are independent AV test labs and reviewers whose results can be used to compare detection/protection rates of AVs (also false positive rates), as well as the performance impact in different scenarios. No amount of personal opinion will change those results.