Hi,
One point I forgot to make last time is that I'm a big fan of prototyping.
I have almost no experience whatsoever in Java Swing, but there are
plenty of people in the R community who do - e.g. Simon Urbanek and the
JGR team. In the past, I have had trouble finding any elegant prototypes
(e.g. "Hello, World" applications) interfacing R with Java Swing.
Similarly, I'd love to see "Hello, World" prototypes for R-Qt etc. Of
course for many of those GUI systems originally developed for Linux or
Unix, there is the question of how reliably you can port them to the most
popular operating system in the world and to MacOS X. I love using GIMP
under Windows, but how many Windows software developers would say that
compiling GTK on Windows (e.g. within an R package) is easy? That's why I
liked wxPython/wxWidgets because it didn't attempt the difficult task of
porting GTK to Windows - it just uses native Windows widgets on Windows
and GTK widgets on Linux.
I don't want to emphasize my interest in prototypes too strongly though -
I still think that there is a lot of work to be done beyond any prototype.
Maybe the successful publication of Rcmdr in the Journal of Statistical
Software will give me hope that the academic community will eventually
feel more comfortable asking for money to employ people to do some
challenging GUI-related software development which may not immediately
promise publications within the core research aims of a particular
department/institute. Or maybe [horrible thought], some business outside
the academic community will provide a very expensive but very good GUI
development system for R.
Best regards,
James