[R-sig-Debian] FW: Warnings

Martin Maechler maechler at stat.math.ethz.ch
Fri Dec 7 15:58:08 CET 2007

>>>>> "GG" == Gorjanc Gregor <Gregor.Gorjanc at bfro.uni-lj.si>
>>>>>     on Thu, 6 Dec 2007 20:56:05 +0100 writes:

    >> Thank you, Gregor,
    >> for the feedback,
    >> >>>>> "GG" == Gorjanc Gregor <Gregor.Gorjanc at bfro.uni-lj.si>
    >> >>>>>     on Thu, 6 Dec 2007 14:46:03 +0100 writes:
    GG> Hmm,
    GG> another one, but error this time!
    li> library(Matrix)
    GG> Loading required package: lattice
    >> >> round(1)
    GG> Error in .Internal(round(x, digits)) : no internal function "round"
    >> "Of course", I cannot reproduce that either.
    >> It must a broken setup / mix of R version and package versions.
    >> Note that 0.99875-2 does *not* work (correctly) with R-2.6.0 and
    >> newer.

    GG> I have a clean/fresh install so this is related to
    GG> package not updated in Ubuntu.  IMHO this indicates a
    GG> "buglet" in Depends syntax. If I am right, 0.99875-2
    GG> should not be installed on R-2.6.0 or newer. Using
    GG> Depends somenthing like R (<= 2.5.1) would result in
    GG> proper dependencies, but I see that this is leading to
    GG> additional work for package developer in general.

It's not possible!

When we released 0.99875-2  ``aeons ago''
we couldn't have known that it would not work with R 2.6.0
which was several months in the future at that time !!!


    >> OTOH, we (Matrix authors) would be very happy if you used the latest
    >> version of Matrix {from CRAN} rather than the Ubuntu/Debian one
    >> which is currently a bit behind (for good "package dependency"
    >> reasons I think, but still):
    >> 0.999375-3 was released early October, and included quite a bit
    >> of work during the summer.

    GG> OK, Thanks!

More information about the R-SIG-Debian mailing list