make check Error: could not get timezone

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

make check Error: could not get timezone

Stephen Berman
I just built the latest R-patched from source (SVN-Revision: 73548, Last
Changed Date: 2017-10-12) and the build completed without a problem but
`make check' errored out:

  running code in 'reg-tests-1d.R' ...make[3]: *** [Makefile.common:100: reg-tests-1d.Rout] Error 1
  make[3]: Leaving directory '/sources/R-patched/tests'
  make[2]: *** [Makefile.common:275: test-Reg] Error 2
  make[2]: Leaving directory '/sources/R-patched/tests'
  make[1]: *** [Makefile.common:165: test-all-basics] Error 1
  make[1]: Leaving directory '/sources/R-patched/tests'
  make: *** [Makefile:239: check] Error 2

The test log reg-tests-1d.Rout.fail ends like this:

  > ## PR#17186 - Sys.timezone() on some Debian-derived platforms
  > (S.t <- Sys.timezone())
  [1] NA
  > if(is.na(S.t) || !nzchar(S.t)) stop("could not get timezone")
  Error: could not get timezone
  Execution halted

Is this a cause for concern and if so, what should I do?  My system is
Linux From Scratch 8.1 (x86_64, linux 4.12.7, glibc 2.26, gcc 7.2.0).

Thanks,
Steve Berman

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: make check Error: could not get timezone

Stephen Berman
On Fri, 13 Oct 2017 22:36:48 +0200 Stephen Berman <[hidden email]> wrote:

> I just built the latest R-patched from source (SVN-Revision: 73548, Last
> Changed Date: 2017-10-12) and the build completed without a problem but
> `make check' errored out:
>
>   running code in 'reg-tests-1d.R' ...make[3]: *** [Makefile.common:100:
> reg-tests-1d.Rout] Error 1
>   make[3]: Leaving directory '/sources/R-patched/tests'
>   make[2]: *** [Makefile.common:275: test-Reg] Error 2
>   make[2]: Leaving directory '/sources/R-patched/tests'
>   make[1]: *** [Makefile.common:165: test-all-basics] Error 1
>   make[1]: Leaving directory '/sources/R-patched/tests'
>   make: *** [Makefile:239: check] Error 2
>
> The test log reg-tests-1d.Rout.fail ends like this:
>
>   > ## PR#17186 - Sys.timezone() on some Debian-derived platforms
>   > (S.t <- Sys.timezone())
>   [1] NA
>   > if(is.na(S.t) || !nzchar(S.t)) stop("could not get timezone")
>   Error: could not get timezone
>   Execution halted
>
> Is this a cause for concern and if so, what should I do?  My system is
> Linux From Scratch 8.1 (x86_64, linux 4.12.7, glibc 2.26, gcc 7.2.0).

In the meantime I installed R and it seems to be functioning correctly.

I reran the tests with `make -k check' and the above test failed again,
but it was the only one.  I looked at the source of Sys.timezone and on
my system as currently configured (TZ is not set, there is no
/etc/timezone file) this function always returns NA, so the test cannot
succeed.  I built and configured my system following the Linux From
Scratch instructions, so (unless I made a mistake somewhere) it seems
this test is irrelevant for that system.  If this conclusion is correct,
maybe the test can be skipped if /etc/timezone does not exist, or at
least not error out.  Or does that mean my system is somehow
misconfigured?

Steve Berman

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: make check Error: could not get timezone

Jeff Newmiller
Off topic. Read the Posting Guide.
--
Sent from my phone. Please excuse my brevity.

On October 14, 2017 11:02:50 AM GMT+01:00, Stephen Berman <[hidden email]> wrote:

>On Fri, 13 Oct 2017 22:36:48 +0200 Stephen Berman
><[hidden email]> wrote:
>
>> I just built the latest R-patched from source (SVN-Revision: 73548,
>Last
>> Changed Date: 2017-10-12) and the build completed without a problem
>but
>> `make check' errored out:
>>
>>   running code in 'reg-tests-1d.R' ...make[3]: ***
>[Makefile.common:100:
>> reg-tests-1d.Rout] Error 1
>>   make[3]: Leaving directory '/sources/R-patched/tests'
>>   make[2]: *** [Makefile.common:275: test-Reg] Error 2
>>   make[2]: Leaving directory '/sources/R-patched/tests'
>>   make[1]: *** [Makefile.common:165: test-all-basics] Error 1
>>   make[1]: Leaving directory '/sources/R-patched/tests'
>>   make: *** [Makefile:239: check] Error 2
>>
>> The test log reg-tests-1d.Rout.fail ends like this:
>>
>>   > ## PR#17186 - Sys.timezone() on some Debian-derived platforms
>>   > (S.t <- Sys.timezone())
>>   [1] NA
>>   > if(is.na(S.t) || !nzchar(S.t)) stop("could not get timezone")
>>   Error: could not get timezone
>>   Execution halted
>>
>> Is this a cause for concern and if so, what should I do?  My system
>is
>> Linux From Scratch 8.1 (x86_64, linux 4.12.7, glibc 2.26, gcc 7.2.0).
>
>In the meantime I installed R and it seems to be functioning correctly.
>
>I reran the tests with `make -k check' and the above test failed again,
>but it was the only one.  I looked at the source of Sys.timezone and on
>my system as currently configured (TZ is not set, there is no
>/etc/timezone file) this function always returns NA, so the test cannot
>succeed.  I built and configured my system following the Linux From
>Scratch instructions, so (unless I made a mistake somewhere) it seems
>this test is irrelevant for that system.  If this conclusion is
>correct,
>maybe the test can be skipped if /etc/timezone does not exist, or at
>least not error out.  Or does that mean my system is somehow
>misconfigured?
>
>Steve Berman
>
>______________________________________________
>[hidden email] mailing list -- To UNSUBSCRIBE and more, see
>https://stat.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide
>http://www.R-project.org/posting-guide.html
>and provide commented, minimal, self-contained, reproducible code.

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: make check Error: could not get timezone

Stephen Berman
On Sat, 14 Oct 2017 11:49:16 +0100 Jeff Newmiller <[hidden email]> wrote:

> Off topic. Read the Posting Guide.

I did do that before posting; the choice seemed to be between R-help
("intended to be comprehensible to people who want to use R to solve
problems but who are not necessarily interested in or knowledgeable
about programming") and R-devel ("for questions and discussion about R
development and programming").  Obviously I need to install R before I
can use it, and in my case I had to build it before I could install it,
and running the tests is recommended, and that's where the issue arose;
I couldn't tell if this was due to a bug in R or something on my end,
i.e., I needed help, so that's why I asked here.  If you think R-devel
is the right address, I'll ask there.

Steve Berman

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: make check Error: could not get timezone

Jeff Newmiller
Your confusion would be alleviated if you read the whole thing. E.g.

Questions likely to prompt discussion unintelligible to non-programmers should rather go to R-devel than R-help. [...] For example, questions involving C, C++, etc. code should go to R-devel.

It is unfortunate that the errors involved in installing R on some platforms involve understanding C, but it is a reality that goes with such broad platform support and parsing through C compiler errors is not what this list is about.

In addition, anything you do with a patched or development version of R bears the risk of incomplete support, discussion of which belongs on R-devel. Patched versions are intended to fix specific bugs, so if you are not immediately inconvenienced by such a bug or are not prepared to help fix or at least troubleshoot problems then you probably should not be installing those versions.
--
Sent from my phone. Please excuse my brevity.

On October 14, 2017 12:07:37 PM GMT+01:00, Stephen Berman <[hidden email]> wrote:

>On Sat, 14 Oct 2017 11:49:16 +0100 Jeff Newmiller
><[hidden email]> wrote:
>
>> Off topic. Read the Posting Guide.
>
>I did do that before posting; the choice seemed to be between R-help
>("intended to be comprehensible to people who want to use R to solve
>problems but who are not necessarily interested in or knowledgeable
>about programming") and R-devel ("for questions and discussion about R
>development and programming").  Obviously I need to install R before I
>can use it, and in my case I had to build it before I could install it,
>and running the tests is recommended, and that's where the issue arose;
>I couldn't tell if this was due to a bug in R or something on my end,
>i.e., I needed help, so that's why I asked here.  If you think R-devel
>is the right address, I'll ask there.
>
>Steve Berman

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.