[Rd] Re: [R] tcltk problem

[Prof Brian Ripley]

There's no problem in R-devel, but this is a configuration issue and 
without the configuration output we can't help, only guess.

Hint: it is tcltk and you need both Tcl _and_ Tk.  As I recall FC2 comes 
with Tcl and not Tk installed by default, so please check Tk is installed 
by looking at the configure output.


On Thu, 23 Dec 2004, Marc Schwartz wrote:

> [MOVED TO R-DEVEL]
>
> On Thu, 2004-12-23 at 12:47 -0500, Liviu M Vladutu wrote:
>> Quoting Marc Schwartz <MSchwartz at MedAnalytics.com>:
>>
>>> On Thu, 2004-12-23 at 11:32 -0500, Liviu M Vladutu wrote:
>>>>
>>>> Hi all,
>>>> I have R Version 2.1.0 installed on a box running Redhat Fedora Core 2.
>>>
>>> Are you really running 2.1.0 (which is an unreleased development
>>> version) or are you running 2.0.1, which is the present released
>>> version?
>>>
>>> What does the banner indicate when you first start R?
>>
>>
>> Here it's the banner:
>> ------------------------------------------------------
>>
>> R : Copyright 2004, The R Foundation for Statistical Computing
>> Version 2.1.0 Under development (unstable) (2004-12-13), ISBN 3-900051-07-0
>>
>> ------------------------------------------------------
>> It writes Version 2.1.0
>
> Then this should have been posted to r-devel, not r-help, since you are
> using an unreleased version.
>
> Note importantly, the word "unstable" in the banner. It is there for a
> reason.
>
>>> How did you install R (compile from source or use one of Martyn's RPMS)?
>>>
>> I installed R from sources and after I downloaded "all" packages (I'm learning R
>> and I want to see all capabilities).
>
> Is there a particular reason for using the development version rather
> than the released version, which is going to be more stable?
>
> If you are just learning R, why complicate that process by using an
> unstable version of it?
>
>>> Also, please run the following command from a console and post the
>>> results back:
>>>
>>> rpm -q tcl tcl-devel
>>>
>> tcl-8.4.5-7
>> tcl-devel-8.4.5-7
>
>>>> When I try:
>>>> library()
>>>> it shows the package tcltk.
>>>>
>>>> But if I try:
>>>> library(tcltk) I get the same error message like in this thread:
>>>> http://tolstoy.newcastle.edu.au/R/help/01c/3418.html
>>>>
>>>> The same, tcl and tk are installed in /usr/lib; if I try:
>>>> capabilities("tcltk")
>>>>
>>>> I get:
>>>> tcltk
>>>> FALSE
>>>
>>> As is indicated in the above post you referenced, this suggests that you
>>> compiled from source, but R did not locate the requisite files during
>>> the configure process.
>>>
>>>> I cant' find tcltk.so
>>>
>>> Run the following command in a console and post the results back:
>>>
>>> locate tcltk.so
>
>>>> Any suggestions please?
>>>> Thx in advance and Merry Xmas!
>>>> Liviu
>>>
>>> Please provide the above information and we can help further.
>>>
>
> Given that you are running the devel version of R, it is entirely
> possible that something is broken here with respect to tcltk and that
> could feasibly change on a daily basis. That is why using the devel
> version is risky unless you have a particular future development related
> need or are planning to contribute bug reports to R Core.
>
> I'll defer further comment to anyone running r-devel, as I am not as of
> yet, other than to say, that I would strongly urge you to remove it and
> run 2.0.1, either from source if you wish or to use the RPM that Martyn
> Plummer has kindly provided.
>
> HTH,
>
> Marc Schwartz
>
> ______________________________________________
> R-devel at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
>

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595