additional problems with online help (C-c C-v)

[Martin Maechler]

{CC'ed to ESS-help ;  it is of more general interest}

    MHe> Running ESS 5.1.13 with Splus 5.

    MHe> I think that there are still some problems with online help.  Suppose
    MHe> I do the following:

    MHe> 1. M-x S+5 to start Splus.
    MHe> 2. C-c C-v pnorm

    MHe> I then get a buffer that says:

    MHe> cat: illegal option -- c
    MHe> Usage: cat [-benrstuv] [-|File ...]

I think this is ONLY true, 
if you first did M-x S (instead of S+5) and THEN
called  C-c C-v pnorm  which gave you the above bug
and THEN did your above two things.

Explanation:  ESS tries to be smart and *RE*-use all existing help buffers
{in order to save the time to re-format them}.
Unfortunately, this also happens for help buffers with an error message in
them, in the above case it *RE*uses the  *help[S](pnorm)* buffer.

WORKAROUND :  Just kill an ESS help buffer when "it is wrong".

The same phenomenon happens, e.g., in R  {http://www.r-project.org/}
if ask for help about lqs  before calling  library(lqs).
This will give you an error "no help ... for lqs".
If you then do library(lqs) and ask for help again,
you *seem* to still get the same error -- however, ESS is only reusing the
old *help*[R](lqs) buffer  --- if you haven't killed it.

    MHe> 3. If I do C-c C-v rnorm, then I get the usual help screen:

{in a new buffer called  *help[S](rnorm)*

    MHe> Normal (Gaussian) Distribution
                                       
    MHe> DESCRIPTION:

    MHe> Density, cumulative probability, quantiles and random generation for
    MHe> the normal (also called Gaussian) distribution.
   
    MHe> USAGE:

    MHe> dnorm(x, mean=0, sd=1)
    MHe> pnorm(q, mean=0, sd=1)
    MHe> qnorm(p, mean=0, sd=1)
    MHe> rnorm(n, mean=0, sd=1)

    MHe> [snip]

    MHe> So, I think that C-c C-v pnorm should also bring this up, but instead
    MHe> it gives an error message.  Oddly enough, C-c C-v pbeta works even
    MHe> though C-c C-v pnorm bombs.
{explanation as before}.

I think ESS should be improved to *NOT* use an *help* buffer at all
when ess-help-... gives an error message...

--> TO DO for the future.

Martin Maechler <maechler at stat.math.ethz.ch>	http://stat.ethz.ch/~maechler/
Seminar fuer Statistik, ETH-Zentrum  LEO D10	Leonhardstr. 27
ETH (Federal Inst. Technology)	8092 Zurich	SWITZERLAND
phone: x-41-1-632-3408		fax: ...-1228			<><
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
ess-help mailing list -- To (un)subscribe, send
subscribe	or	unsubscribe
(in the "body", not the subject !)  To: ess-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._