R 1.6.0 has gone into final feature freeze as of today.
[...] Keeping with the practice of the last two full releases, I took the 20020910 tarball, and the rec'ed packages, last night and made Debian packages for the unstable distribution. I intend to update these weekly before Oct. 1 to ensure that they do not migrate into testing yet as Debian's unstable is the appropriate environment for these packages.
The current version in known to pass the checks for our Linux/x86 systems but has two known problems on Solaris/gcc: An optimizer bug in gcc 3.0.3 causing "cluster" to fail in mona.f and a convergence issue with the glns example in nlme.
The build daemons have already caught up. The page http://buildd.debian.org/build.php?pkg=r-base has links to the full build logs on all arches. As of right now, we already have alpha, arm hppa, ia64, mips, mipsel, powerpc, sparc, s390 With the exception of arm, all run (and pass) "make check". Arm has a long- standing fpu bug which is triggered by one test. I made a small blunder. I had previously used a naming scheme like e.g. 1.4.1.cvsYYYYMMDD for the 1.5.0 pre-release. This time I wanted to signal the 1.6.0 nature of the release and picked 1.6.0.rc.20020910 -- but that sorts higher than 1.6.0, so I will need to use 1.6.0.rel or 1.6.0.release on October 1. I hope you're not to offended by this deviation from the simpler "1.6.0" . Dirk
According to the latest figures, 43% of all signatures are totally worthless. -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel 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-devel-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._