Bughunting Season!

[Thomas Pfeiffer]

On 11.03.2014 13:37, Martin Gräßlin wrote:
> On Tuesday 11 March 2014 09:41:35 Jos Poortvliet wrote:
>> On Monday 10 March 2014 20:10:48 Sebastian Kügler wrote:
>>> On Monday, March 10, 2014 18:11:28 Thomas Pfeiffer wrote:
>>>> On 09.03.2014 14:13, Luca Beltrame wrote:
>>>>> In data domenica 09 marzo 2014 14:00:25, Sebastian Kügler ha scritto:
>>>>>> We'll also open up for feedback from the community shortly, with the
>>>>>> Alpha
>>>>>> release that will be tagged on Thursday (and likely be released the
>>>>>> Tuesday
>>>>>> thereafter, stability permitting).
>>>>>
>>>>> Likewise, for those that weren't on IRC when I said it: my plan is to
>>>>> create a specific forum for Next testers, so that *before* bugs are
>>>>> filed, there's some weeding out of other issues.
>>>>>
>>>>>   From you (Plasma + Kwin) I will however need a list or something
>>>>>   similar
>>>>>
>>>>> of>
>>>>>
>>>>> information that should be provided when having issues (like Kwin
>>>>> requiring
>>>>> supportingInformation, for example). If it's too early for this, no
>>>>> problem: but as the release date comes near I think it will be
>>>>> important
>>>>> to ensure that we can properly handle users (and in turn, this may
>>>>> lead
>>>>> to better reports for you developers, or so I hope ;).
>>>>
>>>> Great idea!
>>>> Maybe this would be a good opportunity to try out a new alpha/beta
>>>> testing process?
>>>> Instead of asking people to directly open bugs, we might ask them to go
>>>> to the forum, see whether a thread about a similar issue already exists
>>>> and if not, create one. Then people in the forum could try to collect
>>>> further information together with the testers and then our team opens
>>>> bugs when they have enough information for a useful report.
>>>> I think this may lead to a much better "signal-to-noise ratio" than
>>>> people flooding bugzilla directly.
>>>> What do you think?
>>>
>>> I think that this would increase overhead dramatically. We'd have to ask
>>> all the same questions that are nicely encoded in bugzilla's forms for the
>>> user, and hunt down more detail information.
>>>
>>> This is actually what the UNCONFIRMED -> NEW distinction in bugzilla is
>>> for.
>>
>> On the other hand, it would be *other people* doing most of the asking while
>> it is usually the developers triaging the bugs, right?
>>
>> Not saying what's better either way and the forums don't work for everybody
>> but perhaps it is worth a shot...
>
> Personally I think that it would be a very valuable addition to get the user
> support out of bugzilla. Out of experience most issues reported in bugzilla
> are user support issues. Those are handled better in forums. Thus I like the
> idea and I have brought up this idea in the past.
>
> Now the question is whether this makes sense for the alpha and there I don't
> think it would help. During the alpha we have more experienced users. In fact
> I'd say we have users who know how to use bugzilla. Also the number of bugs
> will still be higher than the user support issues. Probably we can also assume
> that every user support issue will be a bug in the current state of affairs.

Hm yes, you have a point here about alpha testers being more likely 
advanced users. So this may indeed not be the ideal situation where we 
could try out new user support mechanisms.

> Thus I don't think it's a solution for alpha, but I would love to see this at
> a later state (maybe RC?).

Yes, maybe we could try it out in RC and if it works well, have Plasma 
be the test pilot from there on.