This is the mail archive of the
mailing list for the Cygwin project.
Re: Cygwin R usability
- From: Marco Atzeri <marco dot atzeri at gmail dot com>
- To: cygwin at cygwin dot com
- Date: Mon, 17 Aug 2015 16:12:40 +0200
- Subject: Re: Cygwin R usability
- Authentication-results: sourceware.org; auth=none
- References: <tencent_12266CBA39850A13141D72C3 at qq dot com>
On 17/08/2015 15:41, foehn wrote:
I read that there were issues (mainly with the BLAS library) in the Cygwin port of R (http://r.789695.n4.nabble.com/cygwin-R-2-14-0-build-fail-td4035136.html). And till today the official R site has not been supportive of the Cygwin version of R (https://cran.r-project.org/doc/manuals/r-release/R-admin.html#Cygwin), while there are number of people using this "unofficial" port regardless of the caveats showed in the official R site (http://stackoverflow.com/search?q=R+cygwin). A thread on Stack Overflow was posted asking about this "chaos" (http://stackoverflow.com/questions/31962159/cygwin-r-usability) but nobody seemed to be able to address it fully as they are not Cygwin developers. Therefore I send this mail to the insiders and ask specifically for the current situation of the Cygwin port of R:
R developers have a peculiar view on Cygwin. For them what is not build
by them is unsupported. In this view almost all cygwin software is
-) Has the BLAS issue been addressed in the current Cygwin port of R?
Cygwin build use the integrated R BLAS patched, not Lapack one.
So handling of NaN is as expected by R.
-) Beside the BLAS issue, is there any additional problem? Can it pass the building test?
It pass most of tests. There are some failures on IO formatting, but I
never investigated the root cause.
-) In short, is current Cygwin port of R (v3.1.3, 64-bit) usable in serious sense?
I expect so.
If you find any serious issue, let me know.
Thanks to the efforts for porting R to Cygwin and more thanks in advance to anybody responding to this mail,
Problem reports: http://cygwin.com/problems.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple