But having the dialog you mention as a default would be a big mistake. 99.9% of users wouldn't know what to do, and it would be a pure fluke if they selected the most appropriate action.
Well, I gave you the wrong idea about the dialog, if you think that's true. They certainly made the option to "ignore" seem like the worst of all choices, a scary and dangerous decision. If you ever clicked it, it would further nag you about how that was likely to be incredibly unwise and ask you to confirm that option. Then, on every subsequent scan, it would keep flagging that file anyway, and you'd have to ignore it every time.
Personally, I never treated anti-virus software as software to *clean* viruses. I use them for their virus scanner feature, and if they ever come up positive, it's time to reformat the box and start from scratch, hoping your BIOS is clean. The way I see it, if your system has been compromised, your anti-virus could be compromised. I think clicking, "delete" and getting that nice message on how your system is now clean at the end gives the user a dangerous feeling of false comfort. They're really not that much safer than if they had clicked ignore, they're fairly likely to be just as screwed.
From the sounds of it, this sounds like a delete immediately case. It happens on machines that are known to have the malware, and the TOR client is an old version installed in a specific hidden directory. There is no chance of a false positive.
Yeah, I'm not all up in arms against Microsoft for deleting this particular program, mind you. If anything I said implied that, then I was unclear in how I phrased my thoughts. Microsoft appeared very responsible in dealing with this particular case, down to contacting the Tor developers and making sure there was no legitimate reason why Tor would ever have been installed in that way. Kudos.
What gives me pause is that they have the capability of choosing to delete anything off a box. Because there's no guarantee they're going to be responsible with that tool tomorrow, and the next thing you know, a false positive gets deleted. I don't think such an action should even be legal, without explicit consent.
I moved to the Mac a long time ago...Developers shouldn't delegate the hard decisions to users. They should work out the right thing to do, and do it.
Well, that's certainly the Apple philosophy. I'm not saying that disparagingly, and I recognize the advantages of that philosophy, but I will like to point out that it's a preference, not a universal truth. Since you subscribe to it, you're probably very happy with that move to the Mac. I did the Mac thing myself for many years as a result of Apple switching to x86 compatible machines, and as a result of Mac OS X being UNIX. My latest laptop, however, is not an Apple, precisely because I personally hate that Apple philosophy, and it got in my way much more often than it was ever helpful.
I am a software developer. My philosophy, as a developer and as a user, is that a developer doesn't make decisions ever, regardless of whether they're easy or hard. A developer makes suggestions, when the choice appear obvious, in the form of defaults that can be changed in an advanced menu. If it's a hard decision, either because you're not sure what should be chosen, or because the stakes are high (files are going to get deleted, overwritten, the user will have to log out or reboot, etc.), then you don't even pick a default. You ask the question, and allow the user to set his answer as the default in the future, if he so chooses.
Once again, I'm not trying to tell you my philosophy is right and yours is wrong here, I'm just explaining my own preferences. My philosophy is right for me, and I look to use, buy, and create software that abides by it. This is Windows vs. Mac, KDE vs. Gnome stuff...you always have to trade off control for initial user friendliness, and people draw the line of where the cutoff should be differently.