Re: [libreoffice-design] Re: Impress Mode Toolbar

Le 14/01/2016 18:35, Heiko Tietze a écrit :
> On Thursday, 14 January 2016 11:32:59 CET Michel RENON wrote:
> > You said very important things !
> > That's the core problem : currently, there is no user feedback.
> >
> > The design team is very motivated to create good UI, but we (as I
> > participated in the design team) have the "ivory tower" syndrom :
> > we just talk between 3 to 10 people, mostly some people heavily
> > involved in LibreOffice or software development.
> > We never have the feedback from 'lambda/basic' users.
> > Designing a software is much much more than doing IRC or hangouts and
> > talking about one mockup. It should have "use cases, prototypes, user
> > testing and iterate". (it's not a rant against the design team, it's a
> > constructive critic or an enhancement request)
> > I talked about that (and much more already 3 years ago in my blog [1]
> >
> > Few days ago, I started the idea of "eating our own dog food", that is
> > design team should use LO to create mockups, and more if possible.
> > It would be a first step in having user feedback : some internal
> > feedback, but design team should be in front of real usability problems
> > in LO.
> > As a regular user of LO, I'm often angry because of such details that
> > reduce/kill my productivity in my daily job.
>
> That's not fair. We do user survey for years.

Please note that I was talking about "use cases, prototypes, user 
testing and iterate", not surveys.

While surveys can give some results, the real answer is doing user 
testing on prototypes.

It's based on my daily job : we have some requests from users and we 
have to translate that into technical answers.
With our experience, we know that users ask for wishes, hopes, dreams.
And then their feedback is different when they are in front of a 
prototype or real software.

An example : a customer asked for an EDM (Electronic Document 
Management) system. Obviously, he told us that they need complex 
functionalities (access rights, file sharing...). We just installed a 
prototype with Alfresco and basic configuration. After some weeks, the 
customer said that finally, it's too complex, they don't need an EDM.

Another : a customer asked me to modify a form : add a popup linked to a 
new SQL table and even sent me a mockup ! A friend analyzed the request 
and said "no", then talked with the customer to understand why he wanted 
that. Then we proposed something very different but that fits perfectly 
in their workflow to fill that form. So the customer approved the 
prototype and was happy.

And it's like that nearly every day !

And I try to apply that method to my own internal software :
when I develop it, I think that a new functionality will be ok.
Then, in real use, in hurry conditions, I get annoyed by some details 
I've never imagined while designing. And I have to modify it.


> [...] Our work
> is transparent, democratic, and user-centered. Of course not all ideas are
> published, like the considerations about dogfooding - the posting was rejected

Sad news

> since we do not have any developer imporoving Draw

You can see that my 2 last proposals where made with current Draw. [1][2]
It's not perfect, but I could create them easily.

For more precise mockups, I'll have to get more organized (prepare a 
widget gallery, respect original ratio), but nothing impossible.




[1] https://wiki.documentfoundation.org/Michelr/Impress51_group_buttons
[2] https://wiki.documentfoundation.org/Michelr/Impress51_master_panel


> and want to let sleeping
> dog lie. Actually the same was true for the extended toolbar - this posting
> and many more from the design team could be added here as well.
> And to give this a compromise drive: I understand your posting as an appeal to
> not reduce the effort in our work. I would be the last who objects. Surely we
> can improve on many aspects. For instance in respect to the communication, as
> Sophie said. On the other hand, our latest reports [23,24] had only a few
> comments.

I already said during hangout that work on table styles was missing 
precise use cases. They allow to define precise questions, problems that 
the design proposals have to solve.
Currently, there are no precise questions, so how can designers solve them ?

If I take time to define some use cases, will design team use them ?

Cheers,

Michel

--
To unsubscribe e-mail to: design+unsubscribe@global.libreoffice.org
Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
List archive: http://listarchives.libreoffice.org/global/design/
All messages sent to this list will be publicly archived and cannot be deleted