On Wed, Dec 11, 2013 at 2:24 PM, Andreas Förster wrote:

Everyone has a favorite distro, and they all work

I'm not sure about this - we have definitely found some distributions to be easier to support than others, even just compiling Phenix from source. My advice would be to stick to distributions derived from RedHat (i.e. Fedora, RHEL, CentOS, Scientific Linux) or Ubuntu, simply because we will go out of our way to ensure that the binary Phenix installers work on these. (They also have Coot binaries.) That doesn't mean other distributions are necessarily unsuitable, but the software support may be patchier. I can't remember what CCP4 builds on, but I would be surprised if it doesn't support at least the same OSes. -Nat

At the risk of hijacking the thread, is the other type of "safety" considered? Our university is cutting net access to unencrypted computers, after now-publicized hacks originating from a foreign country. They do not seem to understand or know solutions for encrypting linux machines (they are also banishing XP; all FPLCs may have to come offline!). So, to cut to the chase, does anyone run RHEL/CentOS/Scientific Linux on encrypted disks? I am assuming dm-crypt with LUKS would be the way to go, and I would appreciate to hear about how easy it is to set it up and maintain. Can this be done without wiping clean the system? We are not system admins, and don't want to be. Otherwise, we may be forced to switch entirely to Macs (a Mac OS X Server as our SBGrid server?). Thanks, Engin On 12/12/13, 8:32 AM, David Waterman wrote:

CCP4 builds and tests on a handful of Linux distributions, but the distributed binaries are, I think, built on CentOS 5.9. I agree that RHEL derivatives are "safest" for a crystallography platform.

-- David

On 11 December 2013 23:15, Nathaniel Echols mailto:[email protected]> wrote:

On Wed, Dec 11, 2013 at 2:24 PM, Andreas Förster mailto:[email protected]> wrote:

Everyone has a favorite distro, and they all work

I'm not sure about this - we have definitely found some distributions to be easier to support than others, even just compiling Phenix from source. My advice would be to stick to distributions derived from RedHat (i.e. Fedora, RHEL, CentOS, Scientific Linux) or Ubuntu, simply because we will go out of our way to ensure that the binary Phenix installers work on these. (They also have Coot binaries.) That doesn't mean other distributions are necessarily unsuitable, but the software support may be patchier.

I can't remember what CCP4 builds on, but I would be surprised if it doesn't support at least the same OSes.

-Nat

_______________________________________________ phenixbb mailing list [email protected] mailto:[email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

I have been used both Phenix and CCP4 suite on a variaty of Linux distribution, both Debian- and RedHat-based (Ubuntu, Debian, Fedora, CentOS, openSuse, Scientific Linux) on the basis of different architecture and hardware specification (graphic card, processor, etc.). In my personal opinion, Ubuntu is the "safest" way because of the huge hardware support, so that the great majority of updated hardware available today is recognized out-of-the-box. A great support comes also from the community. However, some libraries lacked in the past for specific purposes (e.g. I think there is an issue for mapslicer in ccp4 or a fortran interpreter so that COMO does not run). I have been using openSuse since the version 12.3 and I found that all work correctly and without user efforts. Also the open source graphic driver for ATI cards now work great (this has been a puzzling trouble for a log time since my laptop has a radeon HD5145 which worked only with Ubuntu and proprietary drivers). PyMol and Coot run very well with open source drivers. I have no experience with RHEL but I highly recommend openSuse for a crystallography workstation. Best wishes Fulvio Saccoccia 2013/12/12 Engin Özkan

Let me clear this: the perfect storm of hacks *and* lost/stolen devices with private patient information appears to have led to changes at our institution. I guess these got our IT and lawyers to get religion on data security. Obviously, encryption won't protect against hacking.

That out of the way, nobody has encrypted linux machines?

Engin

On 12/12/13, 10:00 AM, Engin Özkan wrote:

At the risk of hijacking the thread, is the other type of "safety" considered? Our university is cutting net access to unencrypted computers, after now-publicized hacks originating from a foreign country. They do not seem to understand or know solutions for encrypting linux machines (they are also banishing XP; all FPLCs may have to come offline!).

So, to cut to the chase, does anyone run RHEL/CentOS/Scientific Linux on encrypted disks? I am assuming dm-crypt with LUKS would be the way to go, and I would appreciate to hear about how easy it is to set it up and maintain. Can this be done without wiping clean the system? We are not system admins, and don't want to be. Otherwise, we may be forced to switch entirely to Macs (a Mac OS X Server as our SBGrid server?).

Thanks, Engin

On 12/12/13, 8:32 AM, David Waterman wrote:

CCP4 builds and tests on a handful of Linux distributions, but the distributed binaries are, I think, built on CentOS 5.9. I agree that RHEL derivatives are "safest" for a crystallography platform.

-- David

On 11 December 2013 23:15, Nathaniel Echols wrote:

...

On Wed, Dec 11, 2013 at 2:24 PM, Andreas Förster wrote:

...

Everyone has a favorite distro, and they all work

I'm not sure about this - we have definitely found some distributions to be easier to support than others, even just compiling Phenix from source. My advice would be to stick to distributions derived from RedHat (i.e. Fedora, RHEL, CentOS, Scientific Linux) or Ubuntu, simply because we will go out of our way to ensure that the binary Phenix installers work on these. (They also have Coot binaries.) That doesn't mean other distributions are necessarily unsuitable, but the software support may be patchier.

I can't remember what CCP4 builds on, but I would be surprised if it doesn't support at least the same OSes.

-Nat

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing [email protected]http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing [email protected]http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

-- Fulvio Saccoccia, PhD Dept. of Biochemical Sciences Sapienza University of Rome 00185, Rome (Italy) Phone: +39 06 4991 0556