[R-SIG-Mac] built-in editor underlining

John Maindonald john.maindonald at anu.edu.au
Tue Sep 19 00:09:43 CEST 2006


I mostly use emacs and ESS, unless I want to install
packages.  I would use R.app for teaching, but
setting & changing the font size proved too much of
a challenge.  Especially for occasional users of R.app
such as myself, or for novice users, documentation of
such basic (for some users) features as control of font
size, at least to the extent of noting quirks, is more
important than an extensive range of features.

Three further comments:
1) There is virtue, where it is readily doable, in
providing the same range of features, with similar
mechanisms for setting options, as in the Windows
GUI.  This is not because the Windows interface is
an ideal (though it does seem to me commendably
simple and transparent), but because it reduces
surprises for those who move between the two
interfaces.  Of course if it can be done better, as is
the case for managing & installing packages, so
much the better.  One wants good surprises!

2) There is surely virtue in developing an R interface
and an associated editor along parallel tracks, as is
the case for Tinn-R on Windows.  Tinn-R has some
very nice features, but I find it overly feature-rich.  It
seems to me worth looking at for comparison.
Tinn-R's richness of features will become less of a
problem as its features come to be better documented
and users gain better control over what icons appear.

3) It would be useful to add to the R for Mac FAQ,
or place somewhere else that is suitable, a note on
available editor interfaces under OS X.  Perhaps
there is such information somewhere?

I am most impressed with the system you have set
up for testing and for keeping builds up to date.

John Maindonald             email: john.maindonald at anu.edu.au
phone : +61 2 (6125)3473    fax  : +61 2(6125)5549
Centre for Mathematics & Its Applications, Room 1194,
John Dedman Mathematical Sciences Building (Building 27)
Australian National University, Canberra ACT 0200.


On 19 Sep 2006, at 12:31 AM, Simon Urbanek wrote:

> Jörg,
>
> On Sep 17, 2006, at 2:28 PM, Jörg Beyer wrote:
>
>> Your reply lets me fear I stepped onto your feet, which was not my
>> intention, as I already said.
>
> Not at all - at least to the feet you think ;) - simply I really want
> people to tell us when they think - it's very frustrating when people
> don't tell us when they really think.
>
>
>> Which means that I edit my R code with BBEdit, because it offers
>> -- speedy syntax highlighting
>> -- powerful RegEx search and replace, per file/folder/project
>
> This sure sounds like a worthwhile feature
>
>> -- easier navigation through the code (in my opinion; I tend to  
>> use  relatively large files, and have more than only a few open --  
>> just
>> another question of taste)
>
> Do you have things like "jump to the function definition of the
> function that the cursor is it" is mind, or just shortcuts between
> sections? Again, this could be useful
>
>> -- productive automation features (well, that may sound a bit off-  
>> topic, but is part of my personal coding and code organizing
>> style ... again,  only IMHO)
>
> I'm not quite sure what automation you have in mind... I guess you
> don't mean AT ;)
>
>> -- a working clipping system (which could be better ;)
>
> I'm losing you here - clipping?
>
>> -- direct access to UNIX scripts and "filters", which I make  
>> heavy  use of to
>>    "reengineer" or "refactor" code, or for the simple reoccurring
>> things
>>    that can make life easy.
>
> Some thing like "run this file/select through that perl script"?
> Again, easily implemented and it sounds useful.
>
>> -- syntax folding, which is said to be a nice feature, but I  
>> don't  have that
>>    new version, and so I don't have an own opinion.
>>
>> That's rather specialized, isn't it?
>
> Not really. Sure everyone has his own coding style, so why not
> provide the versatility.
>
>> ... while BBEdit on the other hand doesn't have code completion, of
>> course (one of the most criticized aspects). But, you know, I don't
>> profit that much from code completion at the moment. As a relative
>> beginner, I prefer to have the help pages open in R.app and look up
>> what I need. Gives me a better illusion of control ;-) Hopefully
>> that will change, and then in turn may (or may not) change my
>> working style.
>>
>
> Admittedly I couldn't live especially without argument hints as I got
> so used to it. Still, it's interesting, because I would classify many
> of the features you list as "power-features", definitely not for a
> beginner ;).
>
>>
>> A bit philosophy comes in, too. Why should I expect any developer to
>> reinvent certain specialized functionalities, when other tools
>> already offer
>> them, and at the same time integrate seamlessly with R.app? Isn't
>> that just
>> what we want?
>
> Depends. If you found your perfect editor and it integrates
> seamlessly, sure there' is no point in using the integrated one.
> However, that doesn't mean that we should keep it as plain and basic
> as possible. Especially with features that leverage R itself it can
> be more useful if it has additional features - and this is harder for
> other editors to do, so we should take advantage of it. We have
> started in that direction with JGR and it turned out to be a really
> good idea, especially for students and beginners. But then, why
> shouldn't power-users benefit, too?
>
> Cheers,
> Simon
>
> _______________________________________________
> R-SIG-Mac mailing list
> R-SIG-Mac at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-mac



More information about the R-SIG-Mac mailing list