[ESS] the help system

[Martin Maechler]

>>>>> "KDH" == Kasper Daniel Hansen <khansen at stat.berkeley.edu>
>>>>>     on Mon, 16 Feb 2009 09:46:06 -0800 writes:

    KDH> I am using more and more packages that defines generics for basic  
    KDH> functions. Essentially this means that something like

    KDH> ?substr

    KDH> gets picked up by "base" and another package (in this case  
    KDH> Biostrings). In this case R (from a terminal) usually asks whether I  
    KDH> want base:::substr or Biostrings:::substr, through a simple menu. I  
    KDH> have disabled the tcltk version of the menu by
    KDH> options(menu.graphics = FALSE)
    KDH> in my .Rprofile

    KDH> Now for the ESS question. When I access the help through ESS (as I  
    KDH> almost always do), the help system/ESS seems to get very confused by  
    KDH> the menu selection. The first time I type help(substr) it looks as if  
    KDH> everything hangs. The second time it opens up the menu selection in a  
    KDH> separate buffer, but I cannot figure out how to make a selection. In  
    KDH> the end I end up doing help(substr, package = THE_PACKAGE_I_WANT)  
    KDH> which displays the help page inside the inferior R process and not in  
    KDH> a separate buffer.

    KDH> This is becoming more and more of a problem as more and more functions  
    KDH> are present in several packages - usually because they set up a  
    KDH> generic for a standard base R function.

    KDH> It would be amazing if someone could consider taking a look at this.  
    KDH> Or tell me that I really need to start enjoying tcltk menus.

"you really need to start enjoying tcltk menus"

{that's my current pragmatic solution to the problem you
 mention, and I know well ...}

BTW: We (ESS core) know that there are few (but a positive
     number) of very smart emacs lisp programmers among the readers
     of ESS-help;  we have recently contemplated addressing some
     of you to ask for elisp contributions / programming hands
     in the ESS project, such as to also re-juvenate our team.
     If this rings a bell,  please ring back (not to ESS-help !
     but e.g. to ESS-core at ...) !

    KDH> A related issue is that class?SOME_CLASS gets displayed in the  
    KDH> inferior R process as well.

That's an easy one:

I'm always using  ?SOME_CLASS-class  in ESS

Martin