LikeBack in Amarok

[Leo Franchi]

On Mon, Mar 15, 2010 at 8:45 AM, Myriam Schweingruber
<schweingruber at pharma-traduction.ch> wrote:
> Hi all,
>
> I have already told my disapproval of this in IRC, and will not go
> further on that. Time will show I was right.

Always great to know you're humble and willing to engage in productive
discussion.

>> I recognize that there are potential drawbacks and risks, mainly with
>> fracturing the feedback gathering infrastructure that's already in
>> place, but I believe that LikeBack gathers
>> 1) a different kind of feedback than the existing channels,
>> potentially large quantities of brief context-sensitive one-off
>> comments that are not discussions, and should be considered as a
>> statistic rather than real bug reports;
>
> Well, AFAICS, the feedback is exactly the same as the one we already
> get in bug reports and on the forum and IRC, and, if I judge from the
> mail addresses, it's done by exactly the same people.

You do realize you are judging this after it having been enabled for
almost exactly 24 hours? Good to know you've already predicted and
analyzed the future trajectory of this functionality.

Note that the whole point of development is to move things forward. To
restate an often made point that seems to be missed, (1) reverting is
easy and (2) development is fluid, and features change as a result of
testing and feedback.

>> 2) feedback from a different kind of audience than the existing
>> channels, mainly those who are too unexperienced to do a proper
>> wishlist entry but still run git or beta builds.
>
> See above. If you expect feedback from a different audience, this
> would need to be in final releases, not in betas or git code, since
> the people who actually use the git version are experienced enough to
> give feedback through the existing channels. And sorry, I fail to see
> what is difficult in using a forum, I see feedback there from people
> with close to no experience at all.

This is probably the only valid point in the whole email---that we're
reaching the same audience as is already reporting bugs and talking in
the forums. Yes, there is definitely some overlap. That can't be
avoided. However, claiming that the only people who respond on
likeback are ones who have also responded via bugs or on the forum is
just disingenuous. We can't know exactly who will do what. Maybe it
will turn out to have a lot of overlap in the developer builds, and
maybe not. Maybe the feedback we get in LikeBack is significantly
different from the feedback we get in bugreports or forum posts, or
maybe not. No one is claiming this is the be-all-end-all of feedback.

>
>> So my opinion is that the benefits more than justify the risks.
>> Those who are too lazy to do a proper bug report would do a crappy bug
>> report anyway or none at all, so I don't believe bko has something to
>> lose here.
>>
>> Also, it's simply a *nice* thing to do, any user would enjoy the
>> feeling that he's being asked about his opinion, and I for one enjoy
>> going through the messages, which so far are quite positive.
>> From our users: "I like this feedback reporting feature :)" "I also
>> like these buttons to give comments, nice work!!"
>
> BTW, didn't you say you will orient the feature request to the
> wishlist? As for feedback for statistical reasons: so far it doesn't
> look godd, the dislikes are clearly outnumbering the likes, and the
> feature request don't really belong there.
>
> With 4 devs subscribed to LikeBack feedback and all of them doing
> nothing at all so far (I can only see my comments, am I the only one
> who really tried this out in depth?), this will not go very far...

Maybe the root cause of this issue is that we have different opinions
on what LikeBack is for. Seems to me that it's a way for users to
*give feedback*. I have in fact read all the likeback comments that
have been  posted in the last day. No, I didn't comment. Why? It's
*not* a forum or a bugreport. Users are *not* engaging in a
discussion. They are simply sending quick feedback on what they are
feeling at the moment. That is why it is valuable---to gauge the
experience at run time.

I'm not saying the feedback should be ignored. Some will indeed be
replied to. But your blind metric of "oh, only i've commented
therefore (a) everyone else is ignoring it and (b) they are all
missing the point and (c) this is useless" is not the way to look at
it. it has value beyond immediately replying to users.


> PS. I think there are so many more important things that have to be
> fixed in Amarok, our time could be invested much better. If you need
> feedback, there is plenty of that in #amarok, the forum and bugzilla,
> one just has to look at (and react upon, feedback without reaction is
> pointless, even for statistical reason). Bugzilla can provide very
> nice statistics too, BTW

There are always more things to be fixed. That's why we have release cycles.

leo

-- 
_____________________________________________________________________
leonardo.franchi at tufts.edu         Tufts  University 2010
leo at kdab.com                                 KDAB (USA), LLC
lfranchi at kde.org                             The KDE Project