[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[McIDAS #FLG-294974]: Scientific Linux and McIDAS

Hi Bryan,

re:
> I don't think I have posed this question before, if so, please excuse the 
> duplication.
> 
> Does MUG have any plans to offer a supported version of McIDAS-X for 
> Scientific Linux?

I have no idea if the MUG would officially support McIDAS-X on Scientific Linux.
I can say that a large number of folks in the Unidata community have migrated to
CentOS because it is a "freeware" version of RedHat Enterprise.

re:
> See:
> https://www.scientificlinux.org/
> http://en.wikipedia.org/wiki/Scientific_Linux
> 
> It is essentially a freeware version of Red Hat Enterprise Linux.

I believe that the community of CentOS users is larger than those for
Scientific Linux.  The question I have for you is why you are considering
Scientific Linux over CentOS?

By the way, something that may interest you:

You may or may not know that we (Unidata Program Center) use Fedora Linux
releases almost exclusively in-house (we also run CentOS mainly because
it satisfies the RedHat Enterprise requirement for AWIPS-II which we
are actively testing).  I have been very happy with all of the Fedora
releases up until Fedora 13 and 14.  The reason for my unhappiness now
is that McIDAS-X builds on 64-bit Fedora 13/14 do _not_ work correctly.
Evidently there was some change in the runtime C library that is causing
big problems for 64-bit builds, and I have not figured out exactly what
is going on or how to fix things.  Just in case you ever get the chance
to build McIDAS-X under Fedora 13/14, the thing you should look for
is improper parsing/treatment of REFRESH= clauses in IMGDISP commands.
For some reason, a keyword sequence of something like:

REFRESH='ERASE'

or _anything_ more complicated does not work.  When I turn on verbose
"logging" (DEV=CCC), it is seen that the low level code returns back
a blank value for any/all REFRESH= values.  What is _really_ interesting
is that everything works great for 32-bit builds!?  Also, I can build
McIDAS-X (and -XCD) on either Fedora 13/14 and the code will run correctly
if run from Fedora 12 or previous installations.  This behavior is what
has convinced me that there was some fundamental change in the C runtime
library that is causing the problem.

Just thought you might be interested...

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: FLG-294974
Department: Support McIDAS
Priority: Normal
Status: Closed