About three weeks ago, my students megruder and Julie convinced me to change the Friday 15:30 meeting time for my research group to Wednesday at 15:30. I have roughly the same schedule on Wednesdays as on Fridays, modulo some individual meetings with istari_ala and Andrew, so I figured, "six of one, half a dozen of the other".
Well, not so much, as it turns out: my nice reservation for Nichols 236 (the CIS conference room) turned into a pumpkin when we went from Friday to Wednesday. I asked Mrs. Winfough for alternative rooms, and it was the fast train to Davidspadeville. Nichols 233 (the department library)? No dice. 122 (the big classroom, one of only three we have for lectures) was naturally in use. 127 (the smaller classroom)? Nein, nyet, nanka. So, I called
So it transpired that a couple of Wednesdays ago (on 03 Feb 2005), I requested the reservation, but had to use the Linux lab (Nichols 128) on a one-time basis before we got the room. We met there, trying to plan out the spring, 2005 papers for my reading group and seminar. Naturally, a Tegrity cart is the thing to use for such things (*kof* not *kof*). I wheeled Old Bessie into the lab and was momentarily stymied by the fact that our cart no longer has an Epson projector on it. More accurately, we haven't had a cart with a projector on it since fall, 2002. Did I mention that I am a memoryless fishie?
zengeneral and masaga convince me to turn the cart around and just let everyone squint at the Wacom Cintiq as I type. I cast about for a net connection, and we agree that the quickest solution is to unplug the Cat-5 cable from one of the Linux PCs and temporarily hook the cart into its wall jack, the way I do with the wall jack upstairs in 236 or in 127 when the cable connection we usually use goes on the fritz (about once in every 10 lecture days). After the seminar, Sterling, our department sysadmin, came by and pointed out that there actually was an open cable of the same type that we could use. Filed it away for future reference, I did, as we replaced the connection on the Gentoo box.
Fast forward to yesterday (Wednesday) morning. Jason Li, one of my Ph.D. students and the sysadmin for the KSU Bioinformatics and Ecological Genomics groups, came in to our weekly meeting to tell me that his account had been disabled because he had been the last person to log into the Gentoo system whose Ethernet cable I unplugged. He had been looking for Sterling to explain that not only was it I and not he who did it; we didn't actually reboot a system. I went in person to see Sterling and Jason's account was reactivated in short order. It wasn't a big deal, Sterling explained: rebooting systems in our department is strictly taboo, but the deactivation policy is primarily there to get students to come in so the sysadmins can say "please don't do that". Besides, they don't actually deactivate faculty accounts - figuring that we might have a reason for doing it, I hope; because much as I'd like to be Judge Dredd? Not so much.
Which brings me to my point. With all due respect to our sysadmins, it is IMO a breakdown of communications when the means of notification or summons is to deactivate someone's account. The locking of accounts, however briefly, is a serious matter, and even temporary lockouts can have some negative impact on dissertation research when a grad student can't get the account turned back on right away.1 I do realize that students often don't pay attention to e-mail warnings about policy, or requests to explain themselves. It still seems a measure of the last resort to me to lock someone out of their primary login (i.e., printing, shell, and in some cases, e-mail) in order to generate the "obligation to respond". I also feel that it is just a little too likely (which is to say, not entirely improbable) that detected machine reboots are false positives. This one, which I took full responsibility for, is a case in point.
Another communications issue, and one I think is more far-reaching and important, is how policies are announced. Sterling's explanation was the first I had heard of the policy. While writing this entry I came across this entry in Sterling's blog, in which he mentions announcing it to the CIS students' e-mail list. If it was cc'ed to the faculty, though, I never saw it. I do think such policies should be at least posted somewhere if not discussed with us, because it does affect us - including in such rare cases as our graduate students taking the heat for something we did.
I appreciate the hard work our sysadmins do, and for the record, the current crop has been very nice and helpful in telling me about resources and policies when I ask. (I don't deactivate anything to get their attention, though. ;-))
Oh, and one more thing: it's 20/20 hindsight, but since the want of an Ethernet hookup was the root of this, could the available ways to plug in carts and presentation notebooks be posted online (and maybe on signs in the room)? Since the cables are available only in 127, 128, 233, and 236 as conveniences (and not in N16, for example, where I'm told the proper hookup is the RJ-45 wall jack), I'm thinking this would be good information to have or access at need.
1 This was due in part to the sysadmins being out of the office at the moment when Jason stopped by. Of course, the other part was his wanting to go see the admins in person rather than send an e-mail to
email@example.com. I'm not sure whether he was instructed to do so or just figured that sending a reactivation request from GMail or an alternative non-CIS account would hold less weight than going in person. KDD members, please note for future reference: if your account gets deactivated for whatever reason, go in person, but remember to also send e-mail to
helpand leave a message with whomever is around that you did.