[R] RH5.2 bundle

Peter Dalgaard BSA p.dalgaard at biostat.ku.dk
Tue Jan 12 11:36:30 CET 1999

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

> Hello and Happy New R
> Two points:
> 1 Noting the existence of 0.63.2 as a tgz file on CRAN, but being careful
> or lazy depending on how you want to see it, I also note that the binaries
> for Redhat stop at 0.63.1 on RH 5.1. 

0.63.2 is not even 24hr old yet! RPMs will be made in due time, but
the process is outsourced to people who actually have the relevant
machines and OS installed and know their way around the RPM creation
process. (Martyn Plummer, Nassib Nassar, and David Clayton for
i386/alpha/sparc, to be precise).

> I recently got the RH 5.2 Power Tools where I was pleased to see R 0.62.4
> included, lots of libraries including V&R.  This had been compiled into an
> rpm - does anyone know by whom and could the same (or other) good soul
> both link the RH 5.2 version into CRAN and repeat the exercise for 0.63.2
> on RH5.2.  

I believe this was done by someone at RedHat (RH probably wouldn't
want it any other way for security reasons). I forgot the name but he
was corresponding on the mailing list at the time. Is there a problem
running RH5.1 binaries on RH5.2? I can imagine the opposite, but I
don't see a reason for R not being upwards compatible.

> Compilation for the not-so-expert is always messy - you know you find you
> haven't got the right make, the compiler is wrong, one of the libraries is
> in the wrong place for package A but not for package B etc etc.  I am
> talking about Linux of course!  Guido does such a good job for M$. 

Actually, RH5.1 is my development platform and if I can't do 

make install

R just won't get released... (I do use an upgraded egcs compiler, though)

> 2 Critical mass seems to be approaching/have reached the R product.  Is
> now not the time to consolidate (rather than extend) R and produce a
> comprehensive real book for sale in the bookshops and to decorate our
> shelves? 

Yes. Or several, for different target audiences. Books don't just
happen overnight, though. Several members of the core team are in
there with distinct ambitions of providing some better documentation
than the oddly fragmented set of books on Splus. Frits Leisch has just
started an initiative on turning the R Manual into a real manual as
opposed to a function reference with appendices.

> much more than a user manual. Robert and Ross's R Notes, Kurt's FAQ
Er, Bill and Dave (Venables + Smith), mostly.

> library. I know code modification is seductive but there comes a time in
> any job where the Report just has to be written!

Yep. And the process will also be necessary to set the semantics
straight. Lots of things in there that even the developers have a hard
time understanding some times.

> The advantage of this is (a) it is *much* more convenient for the user,
> (b) it gets some money into the R community, especially if it was sold
> directly over the web, which could help finance the next version (c)  it
> is a serious publication that will get cited with an ISBN number and (d)
> it will encourage the managers of the institutions where all you good
> people work to realise that you are not wasting your time but providing a
> serious and viable product. 

You're touching a couple of sticky issues there. Firstly, there's no
money box for R development. It all runs on existing infrastructure
and the involved parties' research time (and spare time). We could of
course set up a fund to be the beneficiary of the sales of a book
authored by "R.Core" and use the money for travel, etc. I'm not sure
the proceeds would be big enough to cover the administration, though.
Secondly, should one stick to the "freely distributable" way, even if
actual books were produced and would any publisher accept such a
scheme? I agree that ISBN numbers are important for several reasons. 

   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

More information about the R-help mailing list