[Libreoffice-ux-advise] ux-advise advice

[Michael Meeks]

Hi Jean-Francois,

On Sat, 2012-02-18 at 17:22 +0100, Jean-Francois Nifenecker wrote:
> I may be somewhat long here and, hopefully, won't bother and take too 
> much of your (precious) time.

	:-)

> Before I go any further, let me introduce myself

	Great to have you around, sounds like you have some useful skills &
insight.

> In a former part of my professional life, I had been an unofficial 
> part-time programmer (Turbo-Pascal, Object Pascal then Delphi).

	Ah - so you have all the skills you require to jump in and be part of
the solution here. That is good to know. If you go to:

	http://www.libreoffice.org/developers-2/

	You can see how to grab the code. I would personally start digging
around: sw/source/core/layout/paintfrm.cxx or thereabouts; if you read
the git log for that you can see -tons- of work and improvements from
Cedric, and perhaps help propose an improvement in patch form.

> (*) who? Are users listened at? On what criteria? Who are the designers? 
> What do they do with the lists discussions? What are their interactions 
> with the devs?

	I guess, the UI experts (a synonoym for designer and artist in what I
write), are to some degree self selecting; and hackers listen to people
who have done good work in the past and/or provided helpful input and/or
provided beautiful artwork and/or are particularly articulate and
thoughtful I guess.

> Ok, that's neat. What makes someone a "UI expert"?

	cf. the above :-) like being a hacker, it requires some history of
pleasant and mutually useful interaction; that credit and position takes
some time to build up.

> Which page do you refer to (there are plenty ;)?

	Whichever pages that you think are out of date and/or inaccurate :-)

> > 	I didn't read his mail; but I suggest that if a hacker has asked for
> > input here, and there has been no negative feedback, and that decision
> > is taken then he did 100% the right thing, and anything else is just
> > time-wasting fluff. Possibly we'll revisit the decision in future.
> > Certainly if there is concrete code, and someone making the status quo
> > even better then we will.
> 
> Well... I *did* give input (see other thread) but this was quite badly 
> received. 

	So - the time to give input is -before- the work is completed, while it
is in progress, while it is still easy to change things, when the
developer is asking for advice :-) When the code change is complete, and
the work has been done, and there are no shortage of other real bugs to
be addressed (ie. just after release) is a pretty terrible time.

> I was treated as a troll. I fail to understand what I did 
> wrong: was I impolite? harsh? did I call people names? I don't think so.

	Perhaps it is the last snowflake that breaks the bough, and gets the
blame prolly :-) if people try to apply pressure to very busy people,
they will get a rather bad reaction often.

> Well, I do. When a dev gets a consensus here, I'd think he gets to 
> [design] and ask for validation.

	Designers who give constructive feedback to developers are supposed to
lurk here :-) hopefully these are also the leaders on the design list.

>  Or is there a brick wall between the two lists? Otherwise, this would
> mean that design decisions are made here by devs without refering to
> anyone in the Desing team.

	There are plenty of people from the design team here. However, they are
not using the list to discuss next-generation-star-trec futures with a
woolly mob of enthusiasts :-) they discuss very concrete, boring, 'real'
things instead :-) That keeps the bandwidth low enough for developers to
interact on the list too. It is a half-way house between the madness in
both the developers list and the design list :-)

> Thanks ;) But no, it would take ages before I can get to the core and, 
> unfortunately, I haven't time enough. Like most of us here.

	Sure - well therein lies the problem. If no-one cares about restoring
this enough to go and change the code, then ... we stick where we
are :-)

> > 	Next time some sort of change happens in this area, I expect the same
> > thing will happen, asking for advice here.
> 
> Meaning that if I want my ideas to spread, I should stay subscribed and 
> post when a question is asked.

	Of course; if you have valuable design input to give, and by your
experience it sounds as if you do, you're welcome to give input on the
list. But be assured that this is just advice. If the repeated advice is
"don't change anything, I'm familiar and happy with the status quo" ;-)
then that rapidly looses it's attraction; there is a consensus that the
existing UI is in need of some major (incremental) change.

> To come back to the question that brought me in this list, ie the 
> (dreaded) text boundaries, I have a strange feeling that this is a 
> question I can't grasp.

	So - I consider the issue closed, until we have a developer that
actually wants to do something about the situation to improve it.
As/when we do - I (personally) quite liked the idea of toggling the
heavy rendering with the 'show non-printing characters' option.

	But then, I'm not doing the hacking - so this is just a random
thought / trivia of advice ;-)

	All the best,

		Michael.

-- 
michael.meeks at suse.com  <><, Pseudo Engineer, itinerant idiot