[ESS] trace(zapsmall, edit=TRUE) problem under ESS?

[Spencer Graves]

Dear Richard: 

      Thanks for the reply.  With your help, I got it to work. 


      This brings me back to the main question in my first post in this 
thread.  Chambers (2008) Software for Data Analysis (Springer) includes 
the following example on p. 70:  

           trace(zapsmall, edit=TRUE)

      This malfunctioned for me under ESS (XEmacs 21.4.20 with ESS 5.3.6 
and GNU Emacs 22.3.1 with ESS 5.3.8) but worked with Rterm 2.7.2.  John 
noted, "This has nothing to do with the example or, really, with trace.  
... One needs to set the editor to something that works with the GUI 
you're using (many GUI's do this themselves)." 

      Might it be feasible to get changes like you described into a 
future release of ESS so ESS users could get this example to work with 
little or no extra effort?  If yes, can I do something to facilitate 
that (without learning Lisp)? 

   
      Thanks again,
      Spencer Graves

Richard M. Heiberger wrote:
> emacsclient works fine on Windows.  I am using GNU Emacs 22.1.1
>
> Add the following two lines to your site-start.el
>
> (server-start)
> (load-file "c:/emacs/ess-svn/lisp/ess-site.el")  ;; set for your machine
>
>
> Start R with M-x R and type the line
>
> options(STERM='iESS', editor='c:/emacs/emacs-22.1/bin/emacsclient') ## set
> for your machine
>
>
> Now test it
>
> abcd <- function(x) x+4
> abcdefg <- edit(abcd)
>
> In the edit window, change the function and then save it with
> C-x # y
>
> Back in *R* look at abcdefg
>
>
> There is one nuisance glitch, I get the message
> Wrong type argument: number-or-marker-p, nil
>
> It seems to cause no harm so I am ignoring it.
>
>
> History:
>
> At the time we defaulted to gnuclient on windows (emacs 18 or 19 I
> think), emacsclient wasn't available on windows.  Now it is included
> as part of emacs 22 and newer.
>
> Rich
>
>
>
>