Cara you have re-ignited the perennial Mac v PC debate!!! You'll be asking about depositing raw diffraction data next ;)
Cheers Ashley Sent from my iPhone On 23/01/2013, at 10:29 PM, Anastassis Perrakis <[email protected]> wrote: > I am of the opinion that the truth lies somewhere in between ... > > Here are my two cents based on personal experience ... > > For example, I am happy myself using a MacBook Pro, which is sufficient for > all my activities, and has all software and data that I need. > Thus, I am myself on the 'new' paradigm side, having a machine with > "mainframe levels of storage and computing power" (I do not run git, > but time machine in a mac has the bits I need from the git idea - as far as I > know git that is). > > In the department, we have about 20-25 scientists. These people need to > ''maintain" and be proficient in many software suites, many more than > a traditional crystallographer (like me in my PhD time for example) would > need: > vector design software for cloning, databases for keeping track of clones, > sequencing viewing software for their clones, > interfaces for crystallisation and biophysical equipment, analysis suites > like Graphpad/Prism, Origin, Kintek (etc etc) for biophysical experiments > ... > .... and lets not forget SAXS software ... Our experience, is that most of > these people like to use a Windows workstation for these (the choice is free), > others prefer a Mac, thats not my point here. Many of that software also > needs to "maintained" by these people... > Also, for a variety of reasons which have to do with IT "support" > restrictions, the Windows machines > we have for them are miss-configured with ancient versions of the windows > "operating" system, but still Ok for many things, but not for really > straightforward use of CCP4/Phenix ... > > My point here is, that these people are less likely to be keen of the idea to > also install and run ccp4/coot/phenix/buster in their machines > (they use pymol/yassara/chimera though locally since they can copy/paste to > their presentations and papers then). > So, we find it useful to keep an old fashioned setup running in parallel. > Linux boxes, hooked to Zalmans or really > big or double LCDs, in a specific room ... People like these for data > processing, all data of many years back are online, incremental backup is > running etc. > For historical reasons we even run NFS/NIS there (I agree its not a great > choice if one would start now). > > My conclusion and advice for labs or departments that have more than 5-6 > people, and are doing crystallography but not as their > "full-time" business is that besides personal PC/Mac, a common room with a > few relatively powerful machines with nice, big, double, > screens, likely also Stereo, is useful for a few reasons: > > 1. Easier to make sure everybody is using the same software more or less > 2. Same machine to everybody - not the situation that a new student gets a > new machine at year 0, which is redundant by graduation time at +3 years > (...or +5,6,7...) > 3. Mixing of people in the room and ability for people to "look over the > shoulder" of others, the point that my colleague Titia Sixma always favours, > which has indeed proved great for teaching others and learning from others. > 4. Centralised "real" backup, availability of diffraction data on-line with > less "mounts"... > > For these machines, centralised user account information and 'home' sharing > is in my view essential, as it allows to "blindly" choose any of the > machines that is available at a time ... and, being a Mac fun, I think Linmux > is better suited for that purpose, financially and practically... > > These said, it reminds me that we need to update the OS, buy a few new > machines, new LCDs ... argh. > > Sorry of this lecture was outside the scope of the original thread. > > Tassos > > > > On 23 Jan 2013, at 9:54, James Stroud wrote: > >> On Jan 22, 2013, at 11:20 PM, Nat Echols wrote: >>> The real difficulty is integrating Macs into a >>> Linux-centric environment, for example configuring NFS, NIS, etc. >> >> That's because NFS and NIS are antiquities left over from the days of >> mainframes. Distributed file systems and user information databases are >> designed for an environment of many workers and few machines, when the >> typical graphics workstation cost $50,000. These days, we argue whether to >> spend an extra $200 on a $500 computer. We have moved to a new paradigm: >> many workers with many more machines, with each machine having >> essentiallymainframe levels of storage and computing power. In other words, >> instead of NFS, you should run git. >> >> James
