Ian,
I agree with two concerns raised so far:
- communication way too late in the cycle
- opt-out instead of opt-in
I do understand that opt-in renders the feature pretty much useless.
On 04 Jun 2016, at 13:42, Ian Skerrett <[email protected]> wrote:
> We have no interest or plans to profile actual individuals. We are looking
> at aggregate data.
I'm trying to understand why a UUID is necessary when you are looking at
aggregate data. Do you have some sample reports/analysis for sharing? I'm
really unable to connect the pieces of that puzzle. Maybe an example can help
me understand.
Also, have you investigated using an anonymized (hashed version) IP address
that is sent by the clients anyway? Splunk should well be able to handle that.
I also suggest to learn from history:
http://news.softpedia.com/news/Google-Chrome-to-Remove-Unique-ID-137535.shtml
<http://news.softpedia.com/news/Google-Chrome-to-Remove-Unique-ID-137535.shtml>
http://www.theregister.co.uk/2010/03/16/google_chrome_unique_identifier_change/
<http://www.theregister.co.uk/2010/03/16/google_chrome_unique_identifier_change/>
I bet they still have that UUID. But that's a use case I could understand. Even
though I don't think that a UUID is really necessary to confirm successful
installations/updates of a product.
-Gunnar
_______________________________________________
cross-project-issues-dev mailing list
[email protected]
To change your delivery options, retrieve your password, or unsubscribe from
this list, visit
https://dev.eclipse.org/mailman/listinfo/cross-project-issues-dev
