[development] Usability testing

[Boris Mann]

On 22-Nov-05, at 10:23 PM, Dries Buytaert wrote:

>> Ken's comments are very interesting (and ring authentic to me).  I
>> understand and agree with Moshe that this is a problem that is not  
>> being
>> dealt with by Drupal right now as an information architecture  
>> problem.
>> However I also believe that there are solution sets that will deal  
>> with
>> many common problems (like the one's that Ken has outlined).  If  
>> we had
>> another module "layer" and made the system easier for folks to create
>> distros it would go a long way towards improving usability for a  
>> large
>> number of users (IMHO).
>
> Agreed.  While this might not be Drupal's fault, I'm confident this  
> is a common problem.  If there is anything we can do to overcome  
> common problems, I'm all ears.  I, for one, am very much interested  
> in the results of their study, as well as their proposed solution(s).

I also think the answers would be interesting but also believe, much  
like custom themes and Jeff's great post (see http:// 
jeff.viapositiva.net/archives/2005/11/on_the_complexi.html), that the  
solutions will be very site/audience specific.

Ken: at some point, we're going to want to work on a citizen  
journalism install profile...I'm surprised the concept didn't come up  
during the one core/many distributions (or maybe people were just  
waiting for me to harp on this :P ). Adrian's install  
system....actually, wait, that's the whole problem....*Drupal's*  
install system, which Adrian has been the main person working on, can/ 
will support the concept. Adrian and Drumm (his work on update system  
is hand in hand with install) are the two "maintainers/architects"  
currently...I'm sure they wouldn't be opposed to further assistance  
here.

Continuing to look at optimizing core and making choices about what  
goes in/out is a somewhat useful discussion...but needs to be  
tempered by contributions. As evidenced by the pathauto discussion,  
there is hidden complexity between features/need (node IDs needn't be  
displayed) and performance (oops! pathauto may have some additional  
usability/management issues which need work before it makes it into  
core). So: in general, the group of us will act as gatekeepers to  
ensure that core remains stable and lean (perhaps leaner than  
usability might dictate) until code is mature enough. And: who's  
signing up to be the path.module maintainer(s)?

A citizen journalism install profile would in essence have two  
audiences: one being the gardeners/editors of the site, geared toward  
managing users (troll issues etc.), highlighting/organizing/ 
moderating content, and some basic site administration. The other  
audience would be the potential citizen journalists and readers --  
reading/being notified of content (from "main" posts to comments) of  
interest to them, seeing interfaces that bubble-up content (popular,  
highly rated, rated by their friends, etc.), and submitting base  
content types. At the very least, this is a custom front page and an  
explorative interface that hides initial complexity ("I'm just a news  
reader") but lets users easily get to more complexity as needed  
("what is this whole submitting content thing? oh, this is easy...").

Man, this list is just like the old days, before the whole issues-to- 
the-list thing...lots of fun, and I'm excited and energized to be  
talking through some of these ideas.

--
Boris Mann
Vancouver 778-896-2747 San Francisco 415-367-3595
SKYPE borismann
http://www.bryght.com