Dear Tom,
I see that RPMS for R 3.4.0 are not successfully built on Fedora, nor
on RHEL 7.
https://koji.fedoraproject.org/koji/packageinfo?packageID=1230
When I build with mock on Fedora 25 I also get a build failure. R is
not able to establish the time zone inside the chroot and this triggers
a regression tests which is new in R 3.4.0. Specifically, in file
tests/reg-tests-1d.R this test fails:
## PR#17186 - Sys.timezone() on some Debian-derived platforms
(S.t <- Sys.timezone())
if(is.na(S.t) || !nzchar(S.t)) stop("could not get timezone")
## has been NA_character_??in Ubuntu 14.04.5 LTS
The work-around is to set the environment variable TZ to any non-empty
string before running make check, in the %check section of the SPEC
file, e.g.
TZ="Europe/Paris" make check
best regards
Martyn
On Thursday, 11 May 2017 17.37.41 WEST Martyn Plummer wrote:> Dear Tom, > > I see that RPMS for R 3.4.0 are not successfully built on Fedora, nor > on RHEL 7. > > https://koji.fedoraproject.org/koji/packageinfo?packageID=1230 > > When I build with mock on Fedora 25 I also get a build failure. R is > not able to establish the time zone inside the chroot and this triggers > a regression tests which is new in R 3.4.0. Specifically, in file > tests/reg-tests-1d.R this test fails: > > ## PR#17186 - Sys.timezone() on some Debian-derived platforms > (S.t <- Sys.timezone()) > if(is.na(S.t) || !nzchar(S.t)) stop("could not get timezone") > ## has been NA_character_ in Ubuntu 14.04.5 LTS > > > The work-around is to set the environment variable TZ to any non-empty > string before running make check, in the %check section of the SPEC > file, e.g. > > TZ="Europe/Paris" make check > > best regards > MartynThank you Martyn. :-) I am rebuilding R bumping the version since the previous version succeeded to build on el6. BTW Tom, it is probably time to get rid of the %defattr(-, root, root, -) after we are already in the Spring here in the North hemisphere. :-) Regards, -- Jos? Ab?lio
Thanks. I've been traveling almost non stop for the last several months. I appreciate your help here. On May 12, 2017 10:25 AM, "Jos? Ab?lio Matos" <jamatos at fc.up.pt> wrote:> On Thursday, 11 May 2017 17.37.41 WEST Martyn Plummer wrote: > > Dear Tom, > > > > I see that RPMS for R 3.4.0 are not successfully built on Fedora, nor > > on RHEL 7. > > > > https://koji.fedoraproject.org/koji/packageinfo?packageID=1230 > > > > When I build with mock on Fedora 25 I also get a build failure. R is > > not able to establish the time zone inside the chroot and this triggers > > a regression tests which is new in R 3.4.0. Specifically, in file > > tests/reg-tests-1d.R this test fails: > > > > ## PR#17186 - Sys.timezone() on some Debian-derived platforms > > (S.t <- Sys.timezone()) > > if(is.na(S.t) || !nzchar(S.t)) stop("could not get timezone") > > ## has been NA_character_ in Ubuntu 14.04.5 LTS > > > > > > The work-around is to set the environment variable TZ to any non-empty > > string before running make check, in the %check section of the SPEC > > file, e.g. > > > > TZ="Europe/Paris" make check > > > > best regards > > Martyn > > Thank you Martyn. :-) > > I am rebuilding R bumping the version since the previous version succeeded > to > build on el6. > > BTW Tom, it is probably time to get rid of the > %defattr(-, root, root, -) > > after we are already in the Spring here in the North hemisphere. :-) > > Regards, > -- > Jos? Ab?lio > > _______________________________________________ > R-SIG-Fedora mailing list > R-SIG-Fedora at r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-fedora >[[alternative HTML version deleted]]