[Rd] Graphics-Device-Size vs. Window-Size: R design flaw

Jens Oehlschlägel jens.oehlschlaegel@bbdo-interone.de
Wed, 15 Nov 2000 16:47:45 +0100

Dear Prof. Brian Ripley,

thank you for answering.

> Can we step back a bit?  I think we could probably do better
> than we do.
> No one envisaged people resizing graphics windows when the system was
> designed (as far as I can see), and locking aspect ratios (which
> S-PLUS now does) when re-sizing looks sensible to me.  We
> might seriously
> want to consider scrollable canvases on graphics devices, too.
> I'm afraid I don't fully grasp Jens' points, in part because he is
> coming from a different usage experience (mainly Windows, I believe)
> than I am.  And now (the run up to 1.2.0) is not a good time to think
> about such changes.  But if Jens could put together a proposal for the
> wishlist on R-bugs, we can return to it.

I am willing to do that, but let me now explain the problem, which triggered
my conclusions.

BTW, I'm also on linux and under linux I face the same trouble.

I had to optimize a very complex graph which was representing some
underlying data structure.

1) It required DIN-A3 (bigger than screen)
2) It required programming and use of strwidth(), as I needed an automated
way to transform data-to-graph
3) It required interacting based on locator() to change the underlying data
basis and replotting data-to-graph
4) (Thus) It required seeing the whole device during interaction, thus
probably scroll_bars/zooming would not help.
5) It required some fine tuning (text wrapping, font size and other
parameters) ON SCREEN before printing.

E.g. there was text surrounded by boxes, and the text had to fit within
boxes tightly. Relatively to the DIN-A3-device the text was small, but it
had to be that size to get a realistic impression of space needs.

Does my point get clear now?

However, beside this particular problem, I'm convinced that a proper
seperation between device features and window features would be very



r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch