[Testing] test pages in the wiki

[Tim Reilly]

Hi Kim,
Thanks for the reply, we've been playing with the XOs for a few days now
and I have a few thoughts about it in terms of QA.

The difficulty of reporting 406 bugs is that some of them may already be
fixed in 432 or some other new build. The new builds are SO much
different that not testing the new build is often sorta nuts.

More test planning and test structuring will be welcome and useful,
however we can find lots of bugs as is even without large plans. You
don't need to be following a test plan to find many, many bugs or things
that aren't working as intended. This is the gray area which I feel we
are having a bit of difficulty. 

Trac, from my use of it, is dissimilar to bugzilla in that suggestions
and discussion in the comments of a trac bug are discouraged (except for
status updates). In desktop qa at Redhat we're used to posting a
UI-related issue, say some button not being useful or some ui piece
feeling out of place, and get feedback on it from other QA people,
developers, etc. After that a conclusion is reach and the bug is either
fixed, closed, or postponed to be fixed later.

We're on such a tight timeframe that (I feel) we have to start to solve
this part of the problem now, rather than after we map out test plans
for everything.

I'm not sure these are questions or concerns with direct answers, I'm
just trying to open up communication a bit. The way we're used to
working is a bit different then how things seem to run in OLPC land.

Tim

On Thu, 2007-05-24 at 16:39 -0400, Kim Quirk wrote:
> Hi Tim and others,
> For bugs, use Trac. I don't recommend putting in bugs for any release
> after build 406 right now because they are too much in flux (too many
> bugs). There is usually an announcement on devel at laptop.org when there
> is a stable release, and you can see the last good release from the
> front page of the wiki. As we get closer to a release we'll probably
> want to take all the builds.
> 
> What we probably need to focus on right now is test planning. If you
> use build 406 to see how things work today, we can all write some test
> cases that can be executed each time we get a 'stable' build -- and
> maybe figure out the best way to automate some of it. 
> 
> If you want to work on basic Functional tests or Secondary Functional
> tests for each Activity, please go right into the Activity from the
> 'Activities' page on the wiki and add a 'Functional Test' section (I
> started this for the first few activities). 
> 
> Please feel free to suggest better format, etc. 
> I would really like to use something for tracking test cases -- Trac? 
> 
> Thanks,
> Kim
> 
> 
> On 5/24/07, Tim Reilly <treilly at redhat.com> wrote:
>         On Thu, 2007-05-24 at 14:43 -0400, Christopher Blizzard wrote:
>         > Kim has been working on some pages that define some ideas
>         around
>         > testing:
>         >
>         > http://wiki.laptop.org/go/Test_issues
>         
>         Now we know what apps (or, err, acts) to focus on, this is
>         great! Should
>         we start filing things in trac? Talk about what we're running
>         into on
>         olpc? What is the most helpful to you Cambridgians?
>         
>         tim
>         
>         _______________________________________________
>         Testing mailing list
>         Testing at laptop.org
>         http://mailman.laptop.org/listinfo/testing
>