[R-SIG-Mac] Problems loading Rcmdr on a Mac 10.7.2 and a Mac 10.6.8

[Prof Brian Ripley]

On 20/09/2013 12:54, John Fox wrote:
> Dear Brian,
>
> On Fri, 20 Sep 2013 07:49:05 +0100
>   Prof Brian Ripley <ripley at stats.ox.ac.uk> wrote:
>> On 20/09/2013 01:52, John Fox wrote:
>>> Dear Sarah,
>>>
>>>> -----Original Message-----
>>>> From: Sarah Hardy [mailto:sarah.hardy at maine.edu]
>>>> Sent: Thursday, September 19, 2013 8:08 PM
>>>> To: John Fox
>>>> Cc: Simon Urbanek; David Winsemius; r-sig-mac
>>>> Subject: Re: [R-SIG-Mac] Problems loading Rcmdr on a Mac 10.7.2 and a
>>>> Mac 10.6.8
>>>>
>>>> Running the "Software Update" and reinstalling R and Rcmdr worked for
>>>> the student with the 10.7.2. We did not reinstall XQuartz. Now I will
>>>
>>> I'm glad that this worked. Apparently, the student had an old version of
>>> X-Windows that was updated by Software Update. As Simon explained, in this
>>> situation installing XQuartz is irrelevant because the other version of
>>> X-Windows is invoked by R.
>>>
>>>> see if it works for the others. BTW - I have identified at least one
>>>> student with a 10.6.8 who had no problems installing R and Rcmdr
>>>> without installing XQuartz (per my instructions) - am going to check
>>>> with other students tomorrow.
>>>
>>> If the student installed a sufficiently up-to-date X-Windows, that should
>>> work, and apparently did. The X-Windows used needn't be XQuartz. My new
>>> instructions suggest that everyone install XQuartz because that will give
>>> them an up-to-date X-Windows if they don't already have X-windows installed,
>>> and will do no harm if they do have another X-Windows.
>>
>> I am not so sure about that.  It will do no harm running CRAN binary R.  But it does mean that there are multiple versions of X11 things about
>> and that will cause problems for other things (including possibly installing packages from source).
>
> Thanks for pointing out this potential problem.
>
> Oh well -- I hoped to have a simple set of instructions, and having everyone install XQuartz followed what I took to be Simon's suggestion (below).
>
> Almost all of the problems people have with the Rcmdr package are on the Mac OS X platform, and since most of these people are not sophisticated in their computer use, having simple instructions is important.
>
> How about the following?
>
> (1) Run System Update.
>
> (2) Check whether X-Windows is installed. (I can restore my previous instructions about how to check.)
>
> (3) If X-Windows isn't installed, install XQuartz.

I think so.  And I have no problem with installing XQuartz on 10.6 or 
10.7 if Apple X11 is not installed.

Our experience on Solaris (which itself comes with two X11 setups) shows 
how easily it is to get them mixed and how hard it can be to disentangle.

>
> etc.
>
> Best,
>   John
>
> . . .
>
>>>> 	> -----Original Message-----
>>>> 	> From: Simon Urbanek [mailto:simon.urbanek at r-project.org]
>>>>
>>>> 	> Sent: Thursday, September 19, 2013 11:49 AM
>>>> 	> To: John Fox
>>>> 	> Cc: 'David Winsemius'; 'Sarah Hardy'; 'r-sig-mac'
>>>> 	> Subject: Re: [R-SIG-Mac] Problems loading Rcmdr on a Mac 10.7.2
>>>> and a
>>>> 	> Mac 10.6.8
>>>> 	>
>>>>
>>>> 	> Well, the I think it reads far more complicated that in needs to
>>>> be.
>>>> 	> You should have a quickstart section that just says:
>>>> 	>
>>>> 	> * Install R 3.0.1 or later
>>>> 	> * Install XQuartz from http://xquartz.macosforge.org
>>>> 	>
>>>> 	> Done. Current R only supports 10.6.8+ and so does XQuartz so the
>>>> above
>>>> 	> covers everything that is supported right now (I recall having
>>>> that
>>>> 	> discussion earlier...).
>>>> 	>
>>>> 	> Then you can have troubleshooting section which says
>>>> 	>
>>>> 	> a) make sure you installed R with Tcl/Tk (it is the default) -
>>>> if in
>>>> 	> doubt, re-install R from CRAN
>>>> 	> b) use Software Update
>>>> 	>
>>>> 	> Then you can have a technical section with the details you show
>>>> below,
>>>> 	> but 99% of users should not need to read it.
>>>> 	>
>>>> 	> Cheers,
>>>> 	> Simon
>
> . . .
> 	
>


-- 
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