Hi, [Preamble, then example/prototype: ] I was looking at the recent discussion of which toolkit to use for GUIs in R in the future, (ignoring for the moment the SciViews/Visual Basic stuff which looks great!) Dirk suggested that wxWindows/wxWidgets is the emerging consensus. I guess we need some working prototypes to evaluate things further. I would still consider FOX as possibly being easier to work with than wxWindows, e.g. see Zed Shaw's post: https://stat.ethz.ch/pipermail/r-sig-gui/2002-October/000016.html which says: "The disadvantage is that wxWindows is a P-I-G! It's huge, requires tons of libraries, always needs a very complicated linking process which takes me forever to get right on any platform." <SNIP> "FOX is very very well designed, extremely easy to work with, well documented, super small," My experience under Windows NT/2000 has tended to agree with Zed in that wxWindows is not easy to build, especially if you are trying to do it with MinGW's g++ (or even cygwin), rather than Visual C++. Most wxWindows/wxWidgets experts would advise you to use Visual C++, but wouldn't it be nice to be able to build R packages with MinGW or similar? Would going through wxPython or similar simplify compilation issues of wxWindows source code, but still present linking issues? I've had a go at working towards a rough prototype of using the FOX GUI toolkit with R, initially through Python (accessing Tk through Tcl isn't too bad is it?). This of course is not a valid argument for why FOX is better than wxWindows, because you can presumably do the same with wxPython. Anybody? I have a Hello-Worldish example of using FOX from R via RSPython, but there are a few glitches to work through: http://bioinf.wehi.edu.au/folders/james/fxpy/ Unfortunately, Python (not FOX) has MS Visual C++ build issues too, i.e. it is very difficult to build in MinGW: http://uucode.com/texts/python-mingw/python-mingw.html and it is quite strict in that it won't let you build and install Python extensions using MinGW if python itself is built with MS Visual C++ (which it always is). But maybe it's OK to just use MSVC-built binaries of Python and the GUI toolkit and then just use the python bindings for the GUI classes. Wishing I could be at userR, James
[R-gui] FXPy R Gui toolkit?
2 messages · James Wettenhall
12 days later
Hi, A few weeks ago, I posted in regards to experimenting with accessing the wxWidgets GUI toolkit from R by going through RSPython. I have now made some quite significant progress with this and would like to offer a prototype of an R-wxPython package for people to play around with if they feel so inclined. The package and documentation can be downloaded from: http://bioinf.wehi.edu.au/folders/james/wxPython/ It requires RSPython 0.5-3 or later. The Windows build of RSPython on OmegaHat (0.4-0) is too old, so either build it yourself or download from : http://bioinf.wehi.edu.au/folders/james/RSPython/ Also required: Python 2.2.x (http://www.python.org) wxPython 2.5.x for Python 2.2 (http://www.wxPython.org) The wxWidgets C++ source code is not required. Regards, James -------------------------------------------------------------------------- James Wettenhall Tel: (+61 3) 9345 2629 Division of Genetics and Bioinformatics Fax: (+61 3) 9347 0852 The Walter & Eliza Hall Institute E-mail: wettenhall@wehi.edu.au of Medical Research, Mobile: (+61 / 0 ) 438 527 921 1G Royal Parade, Parkville, Vic 3050, Australia http://www.wehi.edu.au