Dear Tim, thank you for your reply.

On Friday, August 17, 2012 12:00:31 AM UTC+2, Tim in Boulder wrote:
>
>  On 8/16/2012 2:31 PM, omoling wrote:
> > Honestly, reading your comments, it seems strange to ask oneself why
> > an app requires a permission you don't see the purpose of.
>
> The problem is that you're asking yourself why an app requires a 
> permission, and YOU (or your users) are the ones trying to figure out why 
> it would need that permission, when neither you nor (and especially) your 
> users are qualified to judge in all cases (notably your comments about how 
> READ_LOG can, and by implication should, be accomplished via other means).
>

Not really. Do you see any kind of users' answers/explanations in 
Apprivacy?..
 

>
> > I never(!) said READ_LOG is unnecessary, read more carefully. 
>
> You did imply that people who used it were being lazy or uncreative in 
> their programming practices. If it CAN be necessary, then why would you 
> insult people who use it? It's this attitude of "I know best" when you 
> clearly don't have the right expertise to judge that really angers people.
>

Tim, I did not imply that. I quote myself: "There are lots of ways to do 
some good debugging without reading log files", I don't see the lazy and 
uncreative parts... do you? And it does neither imply that it might be the 
best or worst way to do debugging.
 

>
> > Also, nobody can deny that the log file might(!) contain private data.
>
> From really badly written apps? Sure. But Google failed to give us the 
> permission that we really need -- read OUR log info only.
>

You got it. I don't think you assume all apps are well written, do you?
 

>
> And I hope you never, ever run applications on Windows, Mac OS X, or iOS, 
> since in all three cases the developers can "spy" on lots of personal data 
> without your knowledge. And what about root exploits in Android -- a real 
> bad player could simply include one of those (hidden, of course) in his 
> app, and use it to bypass permissions entirely. An app with no permissions 
> whatsoever could then be given the green light by your app! How is your app 
> "protecting" people, exactly?
>

I miss the part where I'm saying I'm "protecting" users. I try to inform 
users in a better way that it is done so far. It is up to the users to 
decide which apps they install and which apps they keep installed. Anyway, 
this also is a kind of protection, though information.


> > I would like all apps to be transparent about privacy, but it is not the 
> case! That's why
> > Apprivacy was released.
>
> But every single app in the store ALREADY has a list of permissions that 
> it requires, along with a (sometimes TOO scary, but never erring the other 
> way) description of what it does. This is the list you're evaluating, in 
> fact! So Apprivacy does exactly what the market/play store already does for 
> users, only more so on the "scary" and less on the "reasonable analysis" 
> side of the equation.
>

Exactly. The only difference is that Play does show it in a very short 
form, I like to have more information about the matter.
 

>
> >> Transparency, huh? Do you have a definition of transparency?
> > Sure. Just as a start, explaining clearly why an app requires some
> > permissions and what the app needs the data for... One might argue
> > that it is no guarantee that the app is not evil. True. Still I
> > regard is as the better way to go. One might have to analyse deeply
> > an app to evaluate if the app actually is evil or not, and that is
> > not the purpose of Apprivacy.
>
> What IS the purpose, then, other than amplifying the already-too-scary 
> permission description Google makes people accept before they can install 
> any app already? Do you attempt to look at a privacy policy and see if they 
> make an effort to explain all the permissions the app uses? That would be 
> easy enough -- look at the privacy policy and the description for specific 
> permission names. 
>

We obviously don't agree on this point Tim. I think they are not "scary" 
enough since I believe users do usually not open the details of the list of 
required permissions, and from the short description you don't understand 
much.
 

>
> You say that opening the apps' code and analyzing them isn't "the purpose 
> of Apprivacy." I thought the "purpose" was to help protect a user's 
> privacy? I submit that opening up the apps is the ONLY way to actually 
> accomplish that purpose to any degree beyond what Google already built into 
> Android.
>

Sure. I quote my first post here: "... I released Apprivacy recently with 
the aim of motivating users to be more aware of how installed apps deal 
with their privacy.". This is the aim of Apprivacy.
 

>
> In particular, the "users doubt" feature is a terrible idea where people 
> can crucify an app with ABSOLUTELY NO EVIDENCE -- your own screenshot shows 
> that you questioned EVERNOTE, of all things! It has a HUGE number of 
> features that tie into various corners of Android, so of COURSE it needs a 
> lot of permissions! This is like having people vote for who is likely to be 
> a criminal -- guess who will get the most votes? The one who already HAS 
> the most votes! And your app gives it a red banner just for HAVING 
> permissions, so many will already get a negative vote! Either an app is 
> safe or it isn't; a user's opinion isn't going to affect reality; it just 
> makes it MORE scary. 
>

> You've invented app lynching/witch hunting, bringing mob mentality to app 
> rating, and you expect developers to "deal with it" and not complain? An 
> app is a black box. A user clicking "doubt this app" is them looking at the 
> box and taking a guess, and you're string some apps colored red even though 
> that doesn't mean anything. This isn't DATA, this is polling people to find 
> out how long the king's nose is instead of just MEASURING it. 
>
> The Android permission system is probably a good thing overall, limiting 
> access to apps that actually need it, and making users aware of what apps 
> need what permissions. But its downside is that with transparency come 
> users like you who, now that they have that visibility into what an app 
> needs to do, are being freaked out by the sheer quantity of permissions 
> required by complex apps, and who therefore "doubt" apps from well-known 
> and high visibility companies like Evernote. FWIW I doubt Facebook's 
> integrity and privacy policies, but that is 100% orthogonal to what 
> permissions their app requires; if I trusted Facebook, the fact that it 
> required 20 permissions would be irrelevant. I don't trust Facebook, and 
> their app would have to require no more than Internet permission before I 
> would be happy using it. 
>

In my opinion, these are some more +? on the Facebook app from your side.


> On the other hand, I de facto trust Google with the information on my 
> phone (they developed the OS after all! If they were going to be evil, they 
> could do it without asking permission...), and so their apps asking for 
> more permissions seems completely harmless to me. You previously said you 
> wouldn't even install an update of a Google app when it required more 
> permissions. 
>

Tim, I didn't say that, please read MY posts again.
 

> In another breath said that asking for permissions an app doesn't need yet 
> was wrong, and that users should just accept it when you ask for more 
> permissions later. Which is it? You won't even upgrade a GOOGLE app when it 
> asks for more permissions, but you expect "users" to happily upgrade an app 
> from a developer they've never heard of because it's asking for more? 
>

I don't see the point in not trusting unknown developers. I don't trust 
developers who ask to see private date I don't see the purpose for in their 
app.
 

>
> I still think you need to find a new niche. You're not increasing the 
> privacy of users at all, which is the stated purpose of your app, since all 
> you're doing is parroting the permissions the users already can see, and 
> allowing them to lynch apps randomly. Can you point to a SINGLE thing 
> you're doing that DOES actually increase a user's privacy over what the 
> Google permission system does now? 
>

Tim, again, the purpose is stated in my first post here. 
 

>
> Tim
>

Omar 

-- 
You received this message because you are subscribed to the Google Groups 
"Android Discuss" group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/android-discuss/-/sKdvcHiFLOsJ.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/android-discuss?hl=en.

Reply via email to