Meh - that happens a lot. No one's going to get overly upset about it. If you want to see real arguments search for `parental control`.

Anything?

.dat files. Thumbs.db, autorun.inf and .dat files. Please make it stop.

Seriously from a sysadmins viewpoint the Mac way is far more disruptive than all those windows "extras". IMO it is still a big problem OS X still using resource forks.

20 years ago, the university at which I was Dean of Engineering fussed about Macs although many of us had them. My department even set up its own AppleTalk network so we could use an Apple laser printer. By 10 years ago, there were Macs all over the place. The the School of Architecture and Urban Planning used them exclusively as did the Department of Industrial Engineering in my Faculty. Now, I'm told, they're about 60:40, PCs to Macs and the Faculty IT department is quite capable of helping faculty and students with Macs, PCs, and several flavors of Unix. Over in Architecture, they have trouble with Windows.

Seriously, if you were on the Mac side, you'd be on this side of the argument. I'm not saying that Mac spore doesn't drive you batty. I'm saying that Mac sysadmins have similar arguments about Windows.

Seriously I manage loads of mac servers, Linux servers and windows ! So Unfortunatly I know that the Mac "way" is by far the most disruptive ! When you have large storage and millions of files.....I tell ya having 50% of files forked and 5% corrupted due to being old files that will only work when placed on HFS+ you would feel my pain.

On another note who the hell is gonna run their storage off HFS+ and/or OS X. We are slowly but surely phasing out Mac servers, open directory etc in fact any primary use of X server. Fr small offices no problem but anything above 50 people and that's it you have got to be off your rocker !

Maintenance as the driver for system choices
 

Thanks all.

The theme I see running through these comments is an important one: that maintenance of medium sized private networks is challenging and that attempts to limit the complexity by adopting a single manufacturer's software platform are only partly successful.

First, the idea that maintenance is difficult and expensive in medium-sized networks. Few purchasers or users are aware of the effort required to keep private nets working and even fewer fully factor these costs into their procurement. Are medium-sized networks more vulnerable to maintenance problems? I suspect so. Small nets have good performance characteristics because they are small and (mostly) local and because COTS works well here. Big-iron networks are lifeblood for their owners and so get critical planning and maintenance assets as a matter of course.

Medium-sized nets usually result of growth-by-accretion over years and thus have legacy and modern components and a wide variety of uses and users. They tend to be cost centers for their organizations and so get minimal support until something goes wrong -- usually hitting the response-time ceiling or a security breach. But even these events don't shake out enough resource to address the problem; the 'solution' is nearly always incremental and add-on.

Second, the underlying problem will never be resolved. Instead, technological change will simply make it unimportant. In my own case, it is clear that the university is not capable of actually managing the technology it uses. Nor should it be. The idea that every institution -- even one that claims to be technology oriented -- should own and manage its own networks and platforms is simply not on. The existence of IT shops in hospitals, universities, and most businesses is an anachronism -- they are the cottage industries of the 21st century and can be expected to simply disappear over the next decade. Even the most cursory examination of the costs of maintaining a bespoke IT shop will show that, except where deep security is required, it never pays to do so. We already outsource virtually every aspect of operations at every level and web 2 is making that even easier.

The Professor