Collecting User Statistics

[Jaskaran Singh]

Hi Tor,

On Thursday 09 March 2017 03:29 PM, Tor Lillqvist wrote:
> What would be the usefulness of such information? We don't want to let
> what machine and OS users run at some (future) point in time dictate
> what we do in even more future LibreOffice versions.

Ofcourse, it should not dictate what we do. But having some information
is better than no information. Until now, we're just estimating the
numbers but suppose, the number of XP users comes out to be 20%
(hypothetical, but possible since XP is still quite popular in 3rd world
where free software is also popular), then I guess it would be more
sensible to continue support for it. But that's the Organization's
decision. My point is that having data would help us make informed
decisions. Not that data would dictate us what decisions we make.

Actually, my original intent was to get data which would help our
marketing department in figuring out where we lack and where new avenues
wait for us. And in a similar way for development team as well.
Over time, it could serve as an indicator of LO's popularity based on
the category of users (like Mac Users, high-end/low-end system users)

> Like, if we would put such code into 5.4, and in Spring 2018 it would
> say that 5% of users, let's say a million, still run Windows XP or have
> CPUs without SSE2, would that then mean that those users would
> effectively hold us ransom to keeping support for XP and lack of SSE2
> forever? I don't think we want that. This is a developer-driven project
> and we don't like to keep historical baggage forever.
> 
> Commercial suppliers of LibreOffice support are another thing, of
> course. If a paying customer insists on XP support (or some other silly
> thing), the company does that.
> 
> Just my opinion, of course.
> 
> ---tml
> 

Regards,
Jaskaran Veer Singh
IRC Nick: jvsg