RE: [ccp4bb]: CCP4 and other crystallographic programs on Mac OS X

[<Wayne.Lilyestrom@UCHSC.edu>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***

I was able to get around the compiling problems and porting issues on OS X
by just getting the free distribution of Yellowdoglinux.  Their latest
version takes full advantage of the altivec processing, so it is pretty
fast.  On the Newer macs with the Radeon/Nvidia graphic cards you only get
256 colors for now, but new kernels are available almost daily so(hopefully)
the problem should be fixed soon. Data processing is a breeze regardless
CCP4 & solve/resolve compile quite easily with just the standard
modifications to the config scripts. The linux kernel leaves a lot more ram
available for the programs to use and all the ccp4 programs compiled once I
got the latest libs.  This may be the simplest answer until ccp4 comes out
with an official os x distribution.
  Besides you get bragging rights with your tripple boot system.  ;^)

   --Wayne

P.S.  I tried the other linux distributions and they weren't nearly as
up-to-date.



-----Original Message-----
From: Clemens Vonrhein
To: William Scott
Cc: CCP4 Bulletin Board
Sent: 3/15/02 1:16 AM
Subject: Re: [ccp4bb]: CCP4 and other crystallographic programs on Mac OS X

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***

> Also, despite being able to compile CNS 1.1 with g77, we've been
having 
> major difficulties using the Absoft fortran compiler, which I would 
> really like to do in order to take advantage of our two processors.
Has 
> anyone had success with this?

I haven't done any experiments with CNS and various Linux
compilers. But I did some tests with CCP4 and SHARP using the
following compilers:

  - gcc/g77 2.95.* and 3.0.*: work fine, maybe not the fastest
    executables

  - Lahey F95 6.0 and 6.1: again, no problems here. No speed gain
    versus the GNU compilers. But I like the diagnostics (probably not
    of much interest to end users, but important to developers).

  - Portland Group pgf77/pgf90/pgcc 3.2 and 3.3: seems to handle
    'difficult' code the best and still gives correct results. Faster
    code than the two above.

  - Intel Compiler ifc/icc 5.0 and 6.0beta: fairly stable by now. VERY
    fast code.

  - on Alpha/Linux the Compaq Linux version of their compilers: very
    fast code and the usual Compaq/Digital quality.

I played with Absoft and NAGware evaluation cmopilers and gave up on
both: too much 'Windows/Microsoft'-ish style and options. Couldn't
compile CCP4 and/or SHARP (but didn't try too hard since the others
work fine).

Clemens

-- 

***************************************************************
* Clemens Vonrhein, Ph.D.          vonrhein@GlobalPhasing.com
*
*  Global Phasing Ltd.
*  Sheraton House, Castle Park     Tel: +44-(0)1223-353033
*  Cambridge CB3 0AX, UK           Fax: +44-(0)1223-366889
*--------------------------------------------------------------
* BUSTER Development Group      (http://www.globalphasing.com)
***************************************************************