Hi all, Rainer wrote this: > I like it! And I have some proposals for additions.
Very helpful, thanks. I think I included a bit of almost everything you said when I updated the page. On 29 November 2012 21:23, Michael Meeks <michael.me...@suse.com> wrote: > Hah :-) so - JFYI - when people click on send-feedback, we now have > more details about their systems: the exact version of the software > they're running, the platform, the component (writer, base etc.) and > more. That would need propagating to "file a bug" of course; Sure. Keep in mind, what I did is intended as a mockup, not as a final implementation. I think I left out everything that was harder to do, for Rob. I didn't know we add version strings to the feedback URL now, though. That's definitely nice progress. > I wonder - Mozilla have done a lot of this work before us - can we > re-use their backend infrastructure and share development work on that ? > it'd suck to re-invent all their data analytics / query processing > etc. ? Here's how I think Twitter v/ Mozilla's system stack up: Twitter: * no need to set up a new hardware/etc. (at least for collecting feedback; analysing feedback without hardware might be harder) * verified users => less spam (?) * many people have Twitter accounts already, so not such a high hurdle * possible to follow up with users, creating actual contact between developers/designers/QA'ers/marketeers/... and users * people might expect us to follow up with them, and when we don't they become angry (?) * data becomes Twitter's property not ours * dependent on Twitter's general mood and API * probably hard to annotate tweets with LibO/OS version i.m.o: * need to set up hardware * lots and lots of spam and gibberish * no hurdle but clicking the Send Feedback button * impossible to follow up with users * posts are automatically tagged with LibO/OS version * data is our property The (supposed) ease of use (both to us and the user) and the promise of having less spam make Twitter seem attractive to me, still. > is prolly beyond us ATM, but ... perhaps worth collecting if someone > will do real analytics on it. I personally think the feedback is mostly useful for collecting real-world thumbs-ups/thumbs-downs by region and time.[1] As you said, actual text analysis is hard, especially if you have to take into account that our users speak so many languages. Astron. [1] There might be the psychological effect that people feel heard, too. _______________________________________________ List Name: Libreoffice-qa mailing list Mail address: Libreoffice-qa@lists.freedesktop.org Change settings: http://lists.freedesktop.org/mailman/listinfo/libreoffice-qa Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/ Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette List archive: http://lists.freedesktop.org/archives/libreoffice-qa/
