Il giorno mar, 05/05/2009 alle 18.52 -0500, Rajko M. ha scritto:
On Tuesday 05 May 2009 10:38:19 am Alberto Passalacqua wrote:
Probably the initial idea was a bit too wide...
:-)
Yes it was, and hidden in nice presentation.
Probably. I still think it is what we need, not a bit less, to obtain something.
I'll extract points from presentation to normal wiki text. The part I already did helped me to see that BugBusters would be just a part of total testing.
Also, creating testing infrastructure is the way to attract contributors that have no time nor skills, to develop their own.
No sorry. The stable testing team is for dedicated people with some experience, as clarified in the slides. They have at least to know the basics of Linux, how to report bugs, write tutorials and such. Of course others on the long run can learn, but from who if nobody has that knowledge?
Having plan (and infrastructure) that counts on occasional testers, like Google image labeling counts on occasional contributors. I have, still just gut feeling, that what you can see and try on http://images.google.com/imagelabeler/ is possible with almost anything.
I don't think there is an analogy. There is not a basic infrastructure like Google has. We _need_ people that _stably_ in the community perform testing.
Make lists of tests, very small ones, and offer it randomly, within topic/application. It is not necessary to create duel game style tests, it will help a lot to describe one particular task and give expected result, and ask what happened.
I don't want "random testing", that's what we have, and still it leaves too much room to too many problems. Feel free to bring the idea on though.
Instead of asking expertize and hours of user time, you get detailed tests results with user having to spend 5-10 minutes per test.
Well, hours is what we need. In ten minutes you barely report the bug you find on your way, and it is what we do already. Hours don't need to be contiguous, but we need people that is willing to spend some hour and not ten minutes on testing. I think it is not useful to think it can be done "in ten minutes", because it is unrealistic. Bye, A. -- To unsubscribe, e-mail: opensuse-testing+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-testing+help@opensuse.org