Re: R-beta: Two simple suggestions.

Z. Todd Taylor (zt_taylor@pnl.gov)
Thu, 08 May 1997 07:11:17 -0700


Date: Thu, 08 May 1997 07:11:17 -0700
From: zt_taylor@pnl.gov (Z. Todd Taylor)
Subject: Re: R-beta: Two simple suggestions.
To: r-help@stat.math.ethz.ch

Mike Meyer <mikem@stat.cmu.edu> wrote:

> 1) The .RData file (and S .Data area) are cputype dependent.  When R is used 
> from different types of machines (say, HP and Sun) that share a common 
> filesystem (say afs or dfs) then loading the wrong type of .RData file can 
> lead to mysterious behaviour.   In my environment, undergraduate students are 
> the most likely to log into different cputypes, and they end up being quite 
> confused.   To solve this problem in S I set up an S_WORK environment variable 
> that set the name of the .Data area, via the front end shell script.  For 
> example, on the HP I use .Data.hp.   I would like to see the same facility in 
> R.  if R_WORK is missing, then use the standard name.  If R_WORK exists, then 
> use that name as the file to save the R session in.  (Maybe S_WORK exists in 
> Splus only to please me, but it is useful :-).

I agree this would be useful.  Perhaps more useful would be an
architecture-independent database type.  I once had limited
success rolling my own user-defined database in S that utilized
XDR streams to guarantee cross-platform transferrability.  But
wihtout being privy to the underlying C structures, I could
handle only a limited number of structure types.

In my experience, the I/O overhead for XDR is quite small,
especially compared to the general slowness of the S
parser/interpreter.  If R were to internally XDR-ify everything
before writing it out, the benefits would be substantial for
those of us in mixed-architecture environments.  (No more
data.dump/data.restore, no more maintaining duplicate
databases!)

> 2) Again, along the lines of something that S does that is actually useful.   
> In S you can set the S_FIRST environment variable and have this used as the 
> equivalent of the R .Rprofile file.   Might it be a good idea to allow an 
> R_FIRST environment variable as well.  That way I could set user specific 
> preferences that apply no matter what directory I have working in.

Yes.

--Todd
-- 
Z. Todd Taylor
Pacific Northwest National Laboratory
zt_taylor@pnl.gov
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
r-help 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-help-request@stat.math.ethz.ch
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=