Re: Thunderbird’s Future: the TL;DR Version

Robert Kaiser KaiRo at
Tue Aug 12 01:57:30 UTC 2014

Axel Grude schrieb:
> Am 11/08/2014 um 20:42 schrieb Kent James:
>> <rant>I have been trying for 2 years to get permission to publish the
>> available data on Thunderbird usage without success. Yes we have data,
>> yes Thunderbird usage is growing, but these numbers are only available
>> to staff, and when shared with me privately they are always stated as
>> unshareable except in broad statements. Nobody seems to be able to
>> figure out who has permission to release these numbers publicly. </rant>
> Ok, this one I really don't understand. What's the point in an open
> source product when you can't share usage data?

The concern is and always has been what the press makes of that data. 
Mozilla has been burned pretty toughly when this "ADU" ("Active Daily 
Users") data was made public and showed ~160M for Firefox, while a few 
weeks earlier there had been talks in a press release about how Firefox 
had ~500M users. The result was press articles about how Firefox 
"miraculously lost" 2/3 of its users, just because journalists could not 
distinguish estimated amounts of total users based on market share 
numbers (the 500M number) from an actual count of daily add-on blocklist 
update pings that Mozilla Metrics ended badly naming as those "ADU" numbers.
After that bad press of "losing most of its users", policy was made to 
1) rename ADU to "ADI" ("Active Daily Installations"), though I prefer 
just calling them "(addon) blocklist pings" as that's what they really 
are, and 2) not making those numbers public anywhere to avoid them being 

I agree that it's unfortunate we are in this place, but I also 
understand why we are. We even know that there's some basic flaws in our 
ADI metrics, from not sending those pings in all cases where we probably 
should to losing logs of pings on the server side (we are working on a 
remedy for the latter), and the FHR metrics are much better in answering 
usage (but I think neither Thunderbird nor any other non-Firefox product 
has those).

All that said, being open source and using open processes for 
development doesn't mean that all data that is collected through metrics 
and feedback mechanisms has to be open. For privacy reasons, it's even 
good that not all data sent to us via various mechanisms is public. It's 
something we need to weight carefully.
Still, I think overall ADI summary data should be more easily accessible 
(and when you know where, you can even get to some of it easily, e.g. 
via crash-stats).


More information about the tb-planning mailing list