application-based or usergroup-based testing?

[Henrik Nilsen Omma]

OK, I'm done messing with the wiki page structure now (though some 
cleanup of individual pages clearly remains).

In doing that I started looking a bit more closely at our testing plans 
(which we will be developing over the next few weeks). There seems to be 
two ways to go on this: tests based on user groups and applications. 
Both are probably useful, so I'm wondering if we should go for a hybrid 
solution? Of course we would need to make the rationale and border lines 
clear.

User-group-based testing -- is useful for discovering general issues 
related to installing, booting, logging in and charting which tools can 
be useful for certain groups on the desktop and at the CLI. It is also 
useful for people who are new to a11y in getting an idea of the issues 
involved.

Application-based-testing -- is probably also useful for our key 
accessibility apps like Gnopernicus and GOK. These are our cornerstone 
tools for providing access to Ubuntu and should be tested in a range of 
circumstances with various other applications. It might be easier to 
perform rigorous testing of these assistive apps if we have dedicated 
tests for them.

So a test for a certain user group might conclude that 'Yes, performing 
this task will work for this user, provided GOK works well with 
OpenOffice' and would then deffer to the more detailed GOK+OpenOffice test.

Thougts?

- Henrik