[MITgcm-support] RE: MITgcm on OSX-darwin

Martin Losch mlosch at awi-bremerhaven.de
Tue Oct 28 02:20:46 EST 2003


Hi,

I am glad I am no longer alone (o:

There is one important catch with OS X and the MITgcm. The model  
environment relies on the difference between .F and .f files and the  
standard OS X filesystem HFS does not distiguish between upper and  
lower case! Anyone who wants to run the MITgcm on MacOSX should have a  
partition with UFS (Unix file system), which does distinguish between  
upper and lower case. Also I recommend avoiding directory names with  
spaces in them, e.g. I named one partition (very originally) "HD UFS_1"  
and the old genmake of release 1 breaks because of that (I fixed it  
somehow for, but have to look what I did, basically some shell  
variables did not recognize the root of the current working directory  
correctly). Unfortunately the generic OS X applications don't run on  
UFS file systems, so you'll have to partition you hard disk into two  
partitions minimum (good idea anyway). Alternatively you can edit  
genmake and rename replace all .F with something else, e.g. .for, that  
worked for me, too.

You need the Developer tools (c-compiler, make, cvs etc, be careful the  
latest version is not compatible with fink, which you definitely need),  
alternatively some of this can be found at fink.sourceforge.net, but  
not the c-compiler, I think.

LaTeX: try tetex from fink (pdflatex works nicely), and TexShop, that's  
a very user-friendly latex interface, that uses pdflatex by default  
(need tetex, too), all freeware. Once you have an X-Server (I use the  
binaries provided by Apple, but matlab provides one too), you can  
easily install xdvi (or as part of tetex) via fink. There are generic  
dvi-viewers, but they are not free, share-ware with annoying warning  
messages, etc. I use mostly pdflatex.

Martin

On Monday, October 27, 2003, at 08:26  PM, Chris Hill wrote:

> PS This message (from http://dev.mitgcm.org/pipermail/mitgcm-support/)
> has specific compile options for darwin
>
> http://dev.mitgcm.org/pipermail/mitgcm-support/2003-September/ 
> 001714.htm
> l
>
> Chris
>
>> -----Original Message-----
>> From: Chris Hill [mailto:cnh at mit.edu]
>> Sent: Monday, October 27, 2003 2:23 PM
>> To: 'Daniel Enderton'; 'marshall-admin at techsquare.com';
>> 'mitgcm-support at dev.mitgcm.org'
>> Subject: MITgcm on OSX-darwin
>>
>>
>> Dan,
>>
>>  For help on the model setup you can e-mail
>> mitgcm-support at dev.mitgcm.org.  Below are some pointers on
>> OS-X and Unix/Linux and MITgcm in general.
>>
>> Chris
>>
>>
>>
>>> -----Original Message-----
>>> From: Martin Losch [mailto:mlosch at awi-bremerhaven.de]
>>> Sent: Saturday, September 27, 2003 9:43 AM
>>> To: Chris Hill
>>> Subject: Re: OS X support
>>>
>>>
>>> Chris,
>>>
>>> my experience with my PowerBook is great, Unix and a nice graphical
>>> interface with Word, PowerPoint etc. Graphics are great,
>> etc. They are
>>> just a little slow (only 1GHz-CPU, but still faster than the
>>> EUR1,000,000-SunFire Server).
>>>
>>> I run the model with g77 or Absoft Pro Fortran without any problems
>>> (so far). The Fink project (http://fink.sourceforge.net) provides
>>> a lot of
>>> Unix/Linux tools for OS X with an easy-to-use download-and-install
>>> system: teTeX, XEmacs, g77, X11 with various window managers, all I
>>> need. So I can use my PowerBook in the same way as a Linux-based
>>> PC/Laptop.
>>>
>>> I don't have any personal experience with Matlab (my
>> institute won't
>>> buy it for me, but they spent EUR3000 on this machine, I just don't
>>> understand it), but conflicting reports from Samar (who says that
>>> Matlab runs great on OS X) and a colleague of mine here
>> (who struggles
>>> unsuccessfully with the installation of the license-manager, but
>>> impaired by sea-sickness and Beaufort 9 on the Aurora Australis
>>> somewhere south of 62degS). If you search the web you'll find that
>>> there are some issues with Matlab and OS X. It's basically
>>> not generic
>>> but a Unix version, which needs an X-System and a particular window
>>> manager (provided by Mathworks), so that's a bit of a
>> shame. But ask
>>> Samar (spk at ldeo.columbia.edu). As far as hardware goes, I
>>> would go with
>>> as much memory as you can afford (I have 1GB).
>>>
>>> I think that Apple is great, and certainly an alternative to Linux
>>> (although more expensive).
>>>
>>> Martin
>>>
>>> On Friday, September 26, 2003, at 11:52 PM, Chris Hill wrote:
>>>
>>>> Hi Martin,
>>>>
>>>>  We are thinking of getting a few Powerbooks. Any advice?
>>> We want to
>>>> be able to run the model, Matlab, latex etc... on them.
>>>>
>>>> Thanks,
>>>>
>>>> Chris
>>>>
>>>
>>
>>
>>
>>> -----Original Message-----
>>> From: Daniel Enderton [mailto:enderton at MIT.EDU]
>>> Sent: Monday, October 27, 2003 1:56 PM
>>> To: marshall-admin at techsquare.com
>>> Subject: Re: waiting for computer...
>>>
>>>
>>> Jamie (or others),
>>>
>>> It is a bit hectic getting the computer set up, though a few things
>>> are coming together nicely.  Here are some of my outstanding issues
>>> that I a looking for a bit of help with:
>>>
>>> 1.  How do I print off my new computer?
>>> 2.  Matlab disc, MS office disc
>>>
>>> Also, if anyone has any good ideas for OSX applications for
>> dvi -> pdf
>>> (through ps possibly) or directly from tex ->pdf, please
>> let me know.
>>>
>>> Also, I am trying to compile the MITgcm code, but I do not have
>>> -make- command.  Is there anyone that has put MITgcm on an OSX box,
>>> or am I foraging new territory?
>>>
>>> Thanks,
>>> Daniel
>>>
>>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://dev.mitgcm.org/mailman/listinfo/mitgcm-support
>
Martin Losch // mailto:mlosch at awi-bremerhaven.de
Alfred-Wegener-Institut für Polar- und Meeresforschung
Postfach 120161, 27515 Bremerhaven, Germany
Tel./Fax: ++49(471)4831-1872
http://www.awi-bremerhaven.de/People/show?mlosch





More information about the MITgcm-support mailing list