[Rd] Orphaned ncvar? (PR#9603)

[Robert Chatfield]

 From one view, there is no confusion at all, Peter.

There are persistant difficulties with the binary builds at all sites  
I can try.  It seems that
this build is failing tests.  Possibly it's as simple as a change in  
name for the binary, but
it's impossible for me to guess the correct name.

Of course I tried source builds, but they also present real  
problems.   Techniques for
making libraries known ... and which libraries,    ... .h , ...lib,   
more specific dynamic iibraries ...
way beyond the poor soul who's paid to do statistics and science ...   
well, he is paid if he's lucky
and the funding is not suddenly pulled due to "shortfalls"   ... but  
it is.     ... We all try.

There is some help in what you mentioned ...

> I have difficulties seeing even what it is you're trying to  
> achieve, but for one thing, manipulating .libPaths is NOT the way  
> to help anything find include files and C libraries  
> (LD_LIBRARY_PATH and configure options might be).

So what is the correct way and what are the techniques an humble  
atmospheric scientist
might figure this out based on the documentation that the scientist  
can _readily_ find.  I do try.

For example "Having difficulty with binary builds?    See ... and ...  
and ... ; note that ... documentation
is completely out of date and you had best write  
to ... at ... . ... . .. to find out what we all know,
and .... somehow ... you seem not to know.

(flame off)
best regards, Bob Chatfield
... big fan of R


On Apr 7, 2007, at 12:41 PM, Peter Dalgaard wrote:

> chatfield at clio.arc.nasa.gov wrote:
>> An orphaned package?  anyone in Switzerland know if there's an   
>> alternative?
>> Note the email.
>> I guess CRAN-R should be notified.
>>
>> regards, Bob C
>>
>>
> This appears confused. Where is the bug in R?
>
> I have difficulties seeing even what it is you're trying to  
> achieve, but for one thing, manipulating .libPaths is NOT the way  
> to help anything find include files and C libraries  
> (LD_LIBRARY_PATH and configure options might be). If the issue is  
> that mail to juerg.schmidli bounces, how about looking him up on  
> the net?
>
>    -pd
>>> The original message was received at Fri, 6 Apr 2007 16:32:57 -0700
>>> from vayu.arc.nasa.gov [143.232.122.22]
>>>
>>>    ----- The following addresses had permanent fatal errors -----
>>> <juerg.schmidli at env.ethz.ch>
>>>     (reason: 550-Callout verification failed:)
>>>
>>>    ----- Transcript of session follows -----
>>> ... while talking to phil1.ethz.ch.:
>>>
>>>
>>>>>> DATA
>>>>>>
>>>>>>
>>> <<< 550-Callout verification failed:
>>> <<< 550 550 5.1.1 User unknown
>>> 550 5.1.1 <juerg.schmidli at env.ethz.ch>... User unknown
>>> <<< 503 valid RCPT command must precede DATA
>>> Reporting-MTA: dns; pagent2.arc.nasa.gov
>>> Received-From-MTA: DNS; vayu.arc.nasa.gov
>>> Arrival-Date: Fri, 6 Apr 2007 16:32:57 -0700
>>>
>>> Final-Recipient: RFC822; juerg.schmidli at env.ethz.ch
>>> Action: failed
>>> Status: 5.1.1
>>> Remote-MTA: DNS; phil1.ethz.ch
>>> Diagnostic-Code: SMTP; 550-Callout verification failed:
>>> Last-Attempt-Date: Fri, 6 Apr 2007 16:33:08 -0700
>>>
>>
>>
>>
>>> Could you help? The notes are just too telegraphic.
>>>
>>> I keep getting missing packages  or missing main in netdcdf messages
>>> in my attempts to install.
>>>
>>> Macintosh OS X (10.4)  this happens to be old-chip PowerPC ...   
>>> other install will be new-chip Intel
>>>
>>>
>>> * binaries are listed on all the R mirrors, but they uniformly  
>>> fail  to work
>>>
>>> * I've tried to build these from source, but need netcdf.
>>>
>>> I see this ---
>>>
>>> Note
>>> The NetCDF and the udunits library must be already installed on  
>>> the  system.
>>> Author(s)
>>> Pavel Michna
>>>
>>> but the installs don't work ...at configure stage.
>>>
>>> I have downloaded netcdf using the fink package manner (debian-  
>>> linux like)
>>> which creates files in
>>> /sw/lib
>>> and
>>> /sw/include
>>>
>>> but have not been able to get these recognized by R at compile time,
>>> even after using .libPaths("/sw/include")  for the new library  
>>> and  confirming
>>> it with .libPaths()
>>>
>>> Either they are missing the library or cannot find "main" or   
>>> something.
>>>
>>> If we are to build these ourselves, it would be great to find  
>>> out  just how
>>> to get libraries recognized in the documentation, ... or for  
>>> there  to be binaries.
>>>
>>> If you have notes within the .tgz files, I have not found those.
>>>
>>> Please, help ! ;-)
>>> regards, Bob C
>>>
>>> 	
>>>
>>>
>>> Dr. Robert Chatfield
>>> Earth Sciences, MS 245-5
>>> NASA Ames Research Center
>>> Moffett Field, CA 94035    USA
>>>
>>> Ph: 650-604-5490  FAX 650-604-3625
>>>
>>> http://geo.arc.nasa.gov/sgg/chatfield
>>>
>>>
>>
>> 	[[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel
>>