[Rd] NAMESPACE file generation issue R 2.14.0 on Debian Squeeze

Gabor Grothendieck ggrothendieck at gmail.com
Wed Nov 9 15:23:45 CET 2011


2011/11/9 Uwe Ligges <ligges at statistik.tu-dortmund.de>:
>
>
> On 09.11.2011 13:52, Gabor Grothendieck wrote:
>>
>> 2011/11/8 Uwe Ligges<ligges at statistik.tu-dortmund.de>:
>>>
>>>
>>> On 08.11.2011 19:08, Gabor Grothendieck wrote:
>>>>
>>>> 2011/11/8 Uwe Ligges<ligges at statistik.tu-dortmund.de>:
>>>>>
>>>>>
>>>>> On 08.11.2011 17:56, Gabor Grothendieck wrote:
>>>>>>
>>>>>> 2011/11/8 Uwe Ligges<ligges at statistik.tu-dortmund.de>:
>>>>>>>
>>>>>>>
>>>>>>> On 08.11.2011 17:04, Gabor Grothendieck wrote:
>>>>>>>>
>>>>>>>> 2011/11/8 Uwe Ligges<ligges at statistik.tu-dortmund.de>:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On 08.11.2011 16:31, Gabor Grothendieck wrote:
>>>>>>>>>>
>>>>>>>>>> 2011/11/8 Uwe Ligges<ligges at statistik.tu-dortmund.de>:
>>>>>>>>>>>
>>>>>>>>>>> I think many people like to help, but we cannot:
>>>>>>>>>>>
>>>>>>>>>>> You say you are under R-2.14.0 and whn you R CMD INSTALL a
>>>>>>>>>>> package
>>>>>>>>>>> with
>>>>>>>>>>> that
>>>>>>>>>>> version of R, it does not have a NAMESPACE in the end?
>>>>>>>>>>> Then
>>>>>>>>>>>
>>>>>>>>>>>  - your R installation is broken or
>>>>>>>>>>>  - you are looking into a library where you have old versios of
>>>>>>>>>>> the
>>>>>>>>>>> packages
>>>>>>>>>>> or
>>>>>>>>>>>  - you belive you are using R-2.14.0 but you are actually using
>>>>>>>>>>> an
>>>>>>>>>>> older
>>>>>>>>>>> version.
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> This is even reproduced on CRAN.  All platforms work except one:
>>>>>>>>>>
>>>>>>>>>> http://cran.r-project.org/web/checks/check_results_sqldf.html
>>>>>>>>>
>>>>>>>>> No, that one is completely unrelated (and already solved, but not
>>>>>>>>> yet
>>>>>>>>> synced
>>>>>>>>> to CRAN master) to the original question you have removed in your
>>>>>>>>> reply.
>>>>>>>>>
>>>>>>>>
>>>>>>>> OK.  One would have thought that the checks on CRAN would be
>>>>>>>> consistent with the package pages which link to them.
>>>>>>>
>>>>>>> I only see consistent information on that page.
>>>>>>
>>>>>> If you go to:
>>>>>>
>>>>>> http://cran.r-project.org/web/packages/sqldf/index.html
>>>>>>
>>>>>> then the tar.gz file was created using R-2.14.0 but if you then click
>>>>>> on the  check results  link on the same page it takes you to this:
>>>>>
>>>>>
>>>>> Yes. the tar.gz was created with R-2.14.0.
>>>>>
>>>>>
>>>>>> http://cran.r-project.org/web/checks/check_results_sqldf.html
>>>>>>
>>>>>> On the last link on that page it says ERROR and if click on that it
>>>>>> takes you to the output of the check which reveals that it was run
>>>>>> with R 2.13.2.
>>>>>
>>>>> Yes, ince it is checked with different flavors of R, R-oldrelease,
>>>>> R-release, R-devel. See the first column!
>>>>>
>>>>> A source package can be created with an version of R and checked under
>>>>> another version. There is onlyone source package on CRAN, but - just as
>>>>> an
>>>>> exmaple - binaries for R-2.13.x and R-2.14.x. Of course, the checks are
>>>>> applied with the versions stated on the check page. I think you haven't
>>>>> got
>>>>> the whole point of checking with different versions of R.
>>>>>
>>>>>
>>>>>
>>>>>> The Version column on check_results_sqldf.html page refers to the
>>>>>> package's version, not the R version.  To get the R version you must
>>>>>> know to click on each link.
>>>>>
>>>>> No, no, no, no! See the first column!
>>>>> It definitely states if R-olrelease, R-release, R-pacthed or R-devel is
>>>>> used!
>>>>>
>>>>
>>>> OK. That wasn't clear. It would be better if it actually identified
>>>> the release as R 2.13.2, etc.
>>>
>>> We do not want to change those fields daily: R-devel and R-patched
>>> typically
>>> change from day to day - and then things will really become inconsistent
>>> in
>>> some place.
>>>
>>
>> The inconsistencies are less important if they are obvious but more
>> important if they are not.
>>
>> How about indicating R-2.14.0 but not the (2011-09-30 r57179) part.
>
> Honestly, that (svn revision) is the only part that we do not have on the
> front pages but they are given in the log files.

The R version is not on the package's CRAN page, e.g.
http://cran.r-project.org/web/packages/sqldf/index.html  I was
thinking of something like this for the Package source line:

Package built source:	 sqldf_0.4-3.tar.gz (built with R version 2.14.0
Patched (2011-10-10 r12345)

This seems important since that line truly is not the source but is
the built source.  The actual source is not uploaded to CRAN or shown.
A transformation has been applied to it so rightfully that should be
tracked back via the R version.

This would make it very clear not only which version of the R package
you are dealing with but which version of R built it and that can in
certain circumstances be important for precisely identifying it.  This
could also be done for the .zip file, etc. shown on that page.

If that level of detail is not feasible then this would be next best
(just showing the R x.y.z version):
Package built source:   sqldf_0.4-3.tar.gz (built with R version 2.14.0)


-- 
Statistics & Software Consulting
GKX Group, GKX Associates Inc.
tel: 1-877-GKX-GROUP
email: ggrothendieck at gmail.com



More information about the R-devel mailing list