[Rd] packages built against upcoming releases

[Uwe Ligges]

On 17.01.2010 18:16, Jan Theodore Galkowski wrote:
> It is true RHmm is *not* available in the public central repositories.

It is at the usual location at
CRAN/bin/windows/contrib/2.10/RHmm_1.3.1.zip


> I think it's in Bioconductor.

No on CRAN, and that's why I said CRAN. If it was BioConductor, this was 
the wrong mailing list anyway. Since I am the Windows binary maintainer 
for CRAN, I happen to know we have that package available and the line I 
reported is the build line of that package.



> That said, the "installation process" was
> the usual one from the GUI drop down, after electing repositories.  What
> that does behind the scenes is something I do not know, but I don't see
> how I could know.
 >
> Just letting somebody know.  If you don't want to use the information,
> that's cool with me.

But then please tell us
a) What is your version of R?
b) Which mirror did you choose?
c) You installed by clicking from the Windows GUI?
d) Again, what is a "disconnect"?

Hard to say and debug if you do not tell us what you did and which error 
message you got. Please see the posting guide.

Uwe Ligges






> On Sun, 17 Jan 2010 18:04 +0100, "Uwe Ligges"
> <ligges at statistik.tu-dortmund.de>  wrote:
>>
>>
>> On 17.01.2010 16:33, Jan Theodore Galkowski wrote:
>>> I admire package-builders being proactive and having their facilities
>>> ready for upcoming  R releases.  However, if the publicly released
>>> version of R is 2.10.1 and a package is built for R 2.11.0, users get
>>> the embarrassing notice about a disconnect,
>>
>> What a disconnect?
>>
>>
>>> and cannot do much about it.
>>
>> They should install the version that is intended for R-2.10.x releases,
>> not the one for the development version.
>> The RHmm version in the 2.10 repository in CRAN has
>>
>> Built: R 2.10.1; i386-pc-mingw32; 2010-01-03 15:59:39 UTC; windows
>>
>>
>> hence built with R-2.10.x as required.
>>
>>
>>
>>> If it is tedious to roll out a new package on time, perhaps there's a
>>> possibility for automation here.
>>
>>
>> It is automated and we are currently building for R-2.9.x, R-2.10.x and
>> R-devel (the latter in 32-bit and 64-bit), and you have just to say
>> install.packages() and your R wil grab the right version. I suspect you
>> downloaded something manually from the wrong repository.
>>
>> Uwe Ligges
>>
>>
>>
>>
>>
>>> Thanks.  I encountered this yesterday with RHmm.  I'm running R on
>>> WinXP.
>>>
>
> [snip]
>
> --
>    Jan Theodore Galkowski  (o°)
>
>    607.239.1834 [mobile]
>    617.547.1221 [home]
>    617.444.4995 [work]
>
>   bayesianlogic at acm.org
>   http://www.ekzept.net
>
>
>   "Eppur si muove." --Galilei
>
>