On 5 Jun 2009, at 15:36, Matt Lee wrote:

> On the subject of leaks -- we should err on the side of caution with
> regard to leaked albums. There are sites like http:// 
> diditleak.co.uk/ --
> we should use that data and make up a blacklist of albums, and  
> refuse to
> log tracks from those albums, I think.


Personally, I don't like this idea. If implemented it should be  
optional (opt-in or opt-out I don't care).

If I ask a piece of software to do something, it should do it or fail  
with an error message. I'd be pretty unhappy if a new version of  
syslogd decided that it was going to pick and choose which messages  
it was going to log based on some criteria that I didn't understand,  
didn't care about and couldn't change. If I scrobble something, then  
I want the site to record it.

Also, assuming string matching is done to determine which album is  
being listened to (which is usually the case), it risks false positives.

Better to build a selective-scrobble feature into clients, perhaps  
silencing scrobbles by user-configurable criteria - e.g. I might not  
want my extensive Shawaddywaddy collection to be scrobbled; someone  
else might want to scrobble the music they listen to, but not non- 
musical audio files such as recordings of lectures.

Sketch of an implementation for Rhythmbox: Scrobble Free could check  
for the presence of a "private" flag in the ogginfo/ID3 and refuse to  
scrobble any such tracks. A rhythmbox plugin could be written  
enabling users to switch the flag on and off, and perhaps display a  
track's privacy status as an additional column in the playlist. It  
should be noted that this solution also benefits last.fm users.

Ultimately, we should require as little information as possible, but  
allow users who do want to provide us with the data, to tell us  
everything they want to, so that we can use this information to  
provide a better service for them. The important thing is to provide  
a reasonable level of privacy as a default, after which people can  
choose to opt in and out of things.

Switching off IP logging is a good step in that direction. E-mail  
accounts are another good area to look at. Other than that, I think  
we've already found a pretty good balance.

With opt-outs, I think at some point we need to stop and say, "look,  
you can't opt out of that." If people don't want to share their music  
listening data, then why are they signing up to use a music-listening- 
data-sharing website?

With libre.fm there is of course always the possibility of setting up  
your own private server, secured behind hundreds of layers of  
passwords, and with a clause in your last will and testament that it  
be encased in concrete and dumped at the bottom of the Atlantic in  
the event of your untimely demise. People who want abnormally high  
levels of privacy can always pursue that option.

-- 
Toby A Inkster
<mailto:[email protected]>
<http://tobyinkster.co.uk>



_______________________________________________
Libre-fm mailing list
[email protected]
http://lists.autonomo.us/mailman/listinfo/libre-fm

Reply via email to