On 12/03/2014 01:02 PM, Matthew Miller wrote:
On Wed, Dec 03, 2014 at 12:49:30PM -0500, Máirín Duffy wrote:
> usability tesing! that should definitely be one of the measures.
- Usability is increased as measured by user testing specific to each
Yes! But let me be more specific:
For each flavor you'd probably want a set of key things (I'd suggest
5-7) the user ought to be able to do with that flavor, and run them
through those key things and make sure they are achieved successfully by
most if not all of the participants. E.g., for cloud, one of those tasks
could be, "take this docker container and deploy it. Here's the Fedora
website, here's the container. Go." Or whatever.
So by measuring 'usability' I'm specifically saying:
- Measuring the success rate of users in the target audience for each
flavor being able to perform the most critical / key tasks that flavor
should enable them to do.
Another output besides the success rate is going to be a list of
problems the users uncover, of course; hopefully there are less of them
and they are more trivial than not.
For the best buy-in / usefulness of the test results, the WG / SIGs
should be actively involved in determining those 5-7 key tasks to be
assessed for usability (my hope would be they could come pretty much
straight out of the PRD.)
ps sorry for breaking the thread, my phone does that :(