[R] Advice please ...

[Peter Dalgaard BSA]

John Logsdon <j.logsdon at lancaster.ac.uk> writes:

> Hello R-ists
> 
> Have been lurking on the list for a while, I have noticed recently two
> problems:
> 
> 1	There seem to be problems with installing R on SuSe Linux
> version 5.3.  As the list only gets problems and doesn't log the
> successes, has anyone successfully loaded R on SuSe 5.3?  Is this a
> feature of 0.62.3 only or do the problems apply also to earlier versions.

It's a feature of SuSE only. Apparently it only affects some of the
contributed stuff and compiling with f2c solves it. 

> 2	There was a list of outstanding problems from Jim Lindsey that
> have appeared in 0.62.3 that were not there on 0.62.1.  
> 
> I need to upgrade from RH4.1 Linux and was proposing to use SuSe once the
> YasT English problem is solved.  Upgrading is always quite a
> time-consuming affair, particularly as I will copy /usr/local, /home and
> /etc and rebuild completely.

What's wrong with RH5.1?? You still need to rebuild /usr/local because
of the glibc thing, but most everything else upgrades quite smoothly.

> So my question is, what R to load?  I still have 0.50 on my system and
> need to upgrade, given that I will use SuSe 5.3.  Would you suggest I:
> 
> a	Load 0.62.3, the latest version
> b	Load 0.62.1, which works with Jim's libraries
> c	Wait a little for 0.62.3+x next release which will have these
> fixed.

For each new bug reported by Jim, there are tens of bugs fixed between
the two versions. Most of the items on his list were either his
*opinion* on things that he wants us to do differently, annoying
features that have been there all along, or items introduced for
S-plus compatibility. You don't want the bugs of 0.62.1, believe me.

You'll wait a long time for 0.62.4. Next version will be 0.63,
complete with its own set of nasty bugs brought about by internal
restructuring...

> BTW can I suggest, after the recent RSS News article, that all
> publications should reference the software used.  This should help to
> promote R and to ensure that the valuable support from all the
> institutions is retained since citations indexes would show how much it
> was used. The correct reference for R was on the list a little while ago. 

Actually, the correct reference for R would be the manual for the
relevant version. Unfortunately that's not a publication. As soon as
it get somewhere near complete, I hope that we can make sure that
each version (or at least the major ones) gets an ISBN number, etc.

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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 at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._