[R] Problems in Recommending R

Neil Shephard nshephard at gmail.com
Tue Feb 3 15:05:02 CET 2009

Barry Rowlingson wrote:
> 2009/2/3 Neil Shephard <nshephard at gmail.com>:
>> Again I'd disagree, perhaps the most widely used suite of software has a
>> very simple and clean web-site with few bells and whistles, ditto for one
>> of
>> the most popular text-editors.  I am of course referring to the suite of
>> GNU
>> utilities (http://www.gnu.org/) that make a working GNU/Linux
>> distribution
>> and Emacs (http://www.gnu.org/software/emacs/ ).
>  What?!? Surely the most widely-used suite of software is Microsoft
> Windows, and that has a full-on bells, whistles, activeX,
> silverlight-powered web site.

My apologies I ommitted the 'open-source' caveat that Adam had written and I
quoted in my response.  Thus of all the _open-source_ software packages I
have a strong suspicion that it is the GNU utilities that are the most
widely used (since they are what makes up a funtional GNU/Linux
installation, the Linux part simply refers to the code that forms the kernel
and gets the hardware to communicate).

Barry Rowlingson wrote:
>  I'd say there was a direct relationship
> between website glossiness and amount of usage - more people use
> Notepad than Emacs. In which direction the causality (if any) works is
> an interesting question...

Notepad doesn't have a web-site! (If your assertion is true it is the
perfect vindication of the EU taking M$ to court over bundling IE with their
OS ;-)

Theres probably also a relationship between the glossiness of a website (or
indeed software) and its quality/functionality.  Usage is all well and good,
but if you get the wrong answers out it doesn't matter how many people use
it, they'll all be wrong! (viz. using Excel for statistics).  Its a fine

Barry Rowlingson wrote:
>> I like the R web-site, its clean and simple, present key information
>> prominently (manuals, docs, CRAN, RNew and mailing lists).
>  The open-source community should encourage contributions from beyond
> the world of the coder -- graphic designers, translators, writers and
> so on. Careful contributions from non-coders greatly enhance a
> project.
>  Certainly style should not triumph over content but help to express
> the nature of the content. The R website still has a certain y2k feel
> about it, and although I'm sure we'd agree it would be wrong to make
> it all web 2.0 with rounded corners and a tag cloud, there's nothing
> wrong with refreshing a brand every five or six years.

The issue of revamping the web-site arises regularly on this discussion
list.  A few people have said they're willing to help (in this thread and
others in the past), but little has come to fruition.  Refreshing branding
can work two ways though, sometimes the identity and image that has been
built up over time is lost.

The developers of R have focused on what they are good at, which is
developing R.  I get the impression that they are willing to embrace graphic
designers, translators, writers and so on (with some caveats on how it is to
be managed as pointed out by Friedrich), but no one appears to have stepped
up to the oche yet.


View this message in context: http://www.nabble.com/Problems-in-Recommending-R-tp21783299p21810523.html
Sent from the R help mailing list archive at Nabble.com.

More information about the R-help mailing list