[ESS] Should ESS users use fix()? [was ... emacsclient ...]

[Martin Maechler]

>>>>> "Kasper" == Kasper Daniel Hansen <khansen at stat.berkeley.edu>
>>>>>     on Fri, 4 Nov 2005 18:46:14 -0800 writes:

   ...........

    >> I have now tested this and it works. In addition Seth's suggestion
    >> was spot on. For the non-MacOS users (in case this pops up again on
    >> the list): OS X ships with its own version of Emacs. Most users
    >> install a newer more Mac-integrated Emacs. In my case, having
    >> options(editor = "emacsclient")
    >> in R meant that R picked up the "shipped" version of Emacs, not the
    >> one I wanted. I needed to put
    >> options(editor = "/Applications/Emacs.app/Contents/MacOS/bin/
    >> emacsclient")
    >> for it to work (you might have to change the path depending on which
    >> type of Mac-like Emacs you are using, I am using the "Carbon Emacs")

    Kasper> And the way I actually do this is by putting the following in my .emacs
    Kasper> (setq R-editor "/Applications/Emacs.app/Contents/MacOS/bin/emacsclient")

    Kasper> Btw. when I now do something like fix(print.lm) I get the following  
    Kasper> "error" - I use quotation marks because it seems to work fine...
    Kasper> inferior-ess-send-input: Wrong type argument: number-or-marker- 
    Kasper> p, nil
    Kasper> (this is from the *Messages* buffer)

Just a side note:  I  **NEVER** use  fix()  and don't see why a
``real'' ESS user should ever use fix().   

A bit less extremely put: I think you should work with *.R files
to define all your objects and edit the *.R files, send
functions and other stuff from there to the running *R* process
and hence never use things like fix().

I'm interested to learn your arguments pro "fix()"...

Martin