On 29/04/16 3:44 AM, Brad Myers wrote:
We have found that trying to have any particular visualization that
serves lots of needs is rarely successful, and a better strategy is to
try to understand and then more directly answer the specific questions
that developers have. We discuss this observation and various HCI
methods we apply to improve usability for developers in our new paper
to appear soon:
(a) I am not talking about visualisation at all.
(b) Trying to understand and then answer the specific questions *this*
developer has is what I'm on about.
(c) Thank you for that link, it's generally educational and has given me
to think about, and so have some of the references it points to.
Has the work vindicating what I call the "Incompletely-initialised-object
antipattern" been replicated with a language like Swift or Ada or Smalltalk
where constructors have keyword arguments? Taking the case of
making a Point:
Fully initialised object:
p := Point x: 2 y: 17. "Smalltalk"
p = new Point(x: 2, y: 17); "Swift"
p := Point(x => 2, y => 17); "Ada"
Incompletely initialised object:
p = new Point();
p.x = 2;
p.y = 17;
With Ada and Swift, keyword parameters may be freely ordered,
and may be omitted if that makes sense, so it's not clear to me
what cognitive benefit incomplete initialisation might offer.
If you are interested in this topic and will be at the CHI conference
in 2 weeks, please come to our SIG:
Wish I _could_ be there.
You received this message because you are subscribed to the Google Groups "PPIG
To unsubscribe from this group and stop receiving emails from it, send an email
To post to this group, send an email to firstname.lastname@example.org.
For more options, visit https://groups.google.com/d/optout.