Fedora disconnection?

[Máirín Duffy]

Paul W. Frields wrote:
> Well hush my mouth, this sounds great.  But the key factor is the
> "target" part -- and mostly the studies that I've seen users advocate is
> a limitless problem set, without boundaries.  So I mistakenly brought
> that knee-jerk bias to this conversation, which is my own fault.

Ahh, I didn't mean it that way!! I have to totally agree with what you 
originally said, "There is much more to usability testing than simply 
making lists of what people would like to see."

I think a lot of people have the perception:

1. Do usability tests
2. ???
3. PROFI^H^H^H^H^HThe next ipod!!!!11

I more wanted to highlight with the right focus (a very targeted scope 
for testing; you need to know what finely-grained and well-defined tasks 
that you've identified as the most critical for a specific application) 
usability testing can be something that ANYBODY can do and doesn't need 
to involve a lot of time or $$. However like you hinted at:

1. "Usability tests on Fedora" = $$, time, data not useful since it'll 
reflect on the Fedora of three or four versions ago by the time a study 
of that boundless scope is done (and since the tasks may not have been 
chosen strategically it might give you data on the usability of tasks 
that aren't even critical for your users to be able to complete)

vs.

2. "Usability test to determine how difficult it is for a new Fedora 
user to download, burn, and install Fedora in order to try it out" - 
much, much easier to do. It's got a well-defined scope, is a task that 
is quite strategic and critical for Fedora's success (if people can't 
even download and run it, they would never see all the improvements in 
Fedora itself)

I do apologize for beating a dead horse :( but as someone who has a 
usability and design background, I do get a LOT of questions on why we 
don't just usability test Fedora or usability test Satellite or 
usability test this or that to solve problems, so I think this thread 
maybe hit a nerve. :) I think there are a lot of misconceptions about 
how usability testing works and what you actually get out of what you 
put into it (you get a list of problems, not answers to them). So I 
thought it would be useful to make it more clear why simply 'usability 
testing Fedora' is not really something that can be done.

I think the more key thing is to identify those strategic tasks we want 
Fedora users to do. It would help identify the most important 
development priorities. Even without usability testing, we probably 
already have backlogs of bugs and complaints about those tasks that are 
hidden amongst the body of bugs and complaints about less-important 
tasks...

And I think identifying the strategic tasks flows from what Fedora's 
goals are. Who do we want Fedora to be when it grows up? How important 
are less-technical users vs more savvy developers? How important is 
server capabilities vs desktop capabilities? The answers to these kind 
of questions definitely influence what tasks you want users to be able 
to accomplish most easily. I think projects like the Fedora foundations 
and the marketing plan that was done a while back (I can't seem to find 
it on the wiki now :() are really key for defining this.

~m