Setting sane defaults

[Dario Freddi]

2012/2/17 David Edmundson <david at davidedmundson.co.uk>:
>>
>> First of all I think it's great you cared about this. It's also great
>> you guys had a chat with Bjorn at FOSDEM, and we should not drop this.
>> Defaults are one of the areas where usability matters the most - what
>> about taking this opportunity to review that, and ask Bjorn what he
>> thinks about some stuff we're not 100% convinced on (or maybe even
>> stuff we're convinced about)?
>>
> That was a good meeting. He's offered to come to our next sprint.

I would take this up

> As a self-declared usabilty genius the key part of usability is simply
> to monitor feedback. If 99% of us change a setting, that should be the
> default. If no-one uses it, why does the option exist? There's nothing
> to discuss, it's a change based on facts.
> The best people to listen to are the people who actually use our
> software, which clearly includes us. If someone has an opinion without
> experiencing the software (or at least 'interacting' with a
> screenshot) then their opinion isn't very useful IMHO.

Indeed

> As for usability threads I think it's best if you don't use leading
> questions "I think it should be X do you agree?", but simply a "what
> do you expect to happen when you click the presence applet?" or even
> "what have you got this setting set to"? Otherwise you bias the poll.

True, but consider that the case of dealing with a usability expert is
quite different than interacting with average Joe, which surely would
be quite likely more biased when starting the poll.

Now, while I mostly agree with your reasoning, I can tell from my
experience that sometimes things are not as straightforward - indeed
we might be lucky and they'll be as simple as "oh well, this simply
isn't needed as nobody ever did"; but I am afraid with the increase of
the number of users of our software, the probability of getting
complaints followed by potential reconsideration of stuff we
considered being obvious will be much higher (been there multiple
times).

In these cases an advice from an external usability expert (note the
"external" - never meant to question your title :D) might be indeed
precious. So yes - I agree with you that up to now we *might* not need
that, but given our increase in popularity, I wouldn't underestimate
the need of a thorough analysis of at least some parts of our
components in the short term.

> I test out quite a bit of stuff on my housemates, that seems to work well :)

That definitely justifies your "genius" title :) This is indeed very
good to do, and advise everyone else to do that if you still aren't.

>
> Dave