Gas station without pumps

2010 August 8

Epic Tech Fails

Filed under: Uncategorized — gasstationwithoutpumps @ 12:32
Tags: , ,

The Tempered Radical has issued a call for anecdotes from teachers about “epic tech fails” and how they coped with them. I can’t really compete in the one-upmanship here, as I have never relied much on technology for my teaching, and so have not had “epic” failures traceable to technology.

The number-one technology problem here seems to be power failures.  Both Pacific Gas and Electric and the University of California are responsible for under-investment in infrastructure and inadequate maintenance, so that we get multiple power failures a year on campus.  This does not affect teaching very much (except for night-time classes in winter), as few faculty are so chained to their presentation toys that they can’t switch to whiteboards for a lecture or two.  The most memorable power failure for me was not one that affected my class, but a Parks and Recreation theater performance my son was in when he was seven.  The power failed in the theater (actually, for the entire downtown), and they had to finish the performance by flashlight.  The audience of parents seemed more disturbed by the power problems than the kids or the director.

Power failures have caused a lot of damage to our research programs, though, as computer systems fail (we’ve had some power outages that fried the UPS systems, and many power failures are several hours long, exceeding the battery capacity of most UPS systems) and may take days to get back to full functionality. Some of the web services I run require a day of computation for each job submitted, and all work is lost if there is a power failure or a file server failure. The switchover to emergency power for the freezers and other critical equipment in bio labs does not always go smoothly (one incident resulted in losing about a week’s work for a large lab and about $50,000 in lost materials, because the campus electricians had installed the equipment on the wrong circuit—not one that switched over to the emergency generator).

We also have major file server failures just about any time our tech staff tries to do anything major.  Everyone in the School of Engineering lost a full day’s work last Thursday, as the new file server for all home directories failed on its first day, and it took the entire day (6.am. to after 7 p.m.) for the tech staff and the vendor to get it functional.  There are still some things not working as a result of that failure (like they never restarted the procmail demons that some of us use for filtering spam).

I have seen some presentation failures by others, of course.  The most common one is from people who want to show videos from within Powerpoint.  About half the time it seems that they don’t have the correct codec installed or some other glitch causes the video not to work.  About 10% of presenters can’t get PowerPoint to talk properly with projectors (it seems touchier than any other software, which is strange for software supposedly intended for presentations).  Personally, I never use Powerpoint—when I have to use computer projection, I prepare my talk using LaTeX and present it as a PDF file using Acrobat Reader (with Preview as an alternative).

Perhaps the most memorable failure I’ve seen was in an interview talk for a faculty recruit (whose name, mercifully, I don’t remember).  He was working at a national lab and had a decent research reputation, but in his talk he ended up talking to just one member of the audience—the person whose work was closest to his own—so that no one else in the audience had any idea what he was talking about.  He directed his remarks to the one person by name, and never looked at anyone else.  The tech failure was that all his slides ended up being projected sideways.  I didn’t even know that was possible with computer projection, though it was a common problem in the 35mm slide days.  (The faculty recruiting committee agreed unanimously that the recruit was well situated in a national lab, and should not consider a position that required him to teach.)

The tech warnings I give my students include the video codec problem, making sure all fonts used in the presentation are included in the presentation file (at least if there is a chance the presentation will have to be given from a different computer), not using a laser pointer if there is the slightest chance of hand tremor, testing both the computer and the projector ahead of time, avoiding visuals that rely on hue contrast (instead of intensity contrast), and always being able to give their talk even if all visuals fail, rather than fussing with the technology in front of the audience.  I also remind them to sleep the night before and to eat breakfast before a big presentation: I remember one presentation at an international conference where the speaker was running on adrenalin after a 24-hour flight—he got to the end of his presentation and as the first question was asked by the audience he passed out and fell off a 3-foot high stage.

I have had one class damaged by technology, when I was talked into giving a class with some of the students being 1000 miles away.  I had to give lectures in a badly equipped room with poor lighting and flimsy portable whiteboards that did not have enough room for me to move around in front of the whiteboards without tripping over cables.  I spent a lot of time adapting presentations so that the people participating by camera could be included, but it was at a pretty high cost to those who were physically present. As it turned out, only one of the distant students showed up for class on a regular basis, and none signed up for credit. I declined subsequent “invitations” to participate in the program.

1 Comment »

  1. Had a power failure yesterday for about 5 minutes, which happened to take down a computer I had been running a job on for about 50 hours. It took a few hours for everything to be restored to functionality, because one of the file servers that did not go down during the crash needed to be rebooted before any of the compute machines could see it again. Obviously the computers needed to be on UPS power, but they’re not my machines (I haven’t had funding to get new computers in so long that I can’t run the software on any of my own machines—they’re all 32-bit machines, and the software is only available as 64-bit executable).

    I was able to continue the job from a checkpoint, but I think that the checkpointing mechanism was faulty, since the job did not produce answers as good as I had expected. Now I have to decide whether to rerun the entire job for another 50 hours.

    Comment by gasstationwithoutpumps — 2010 August 25 @ 12:36 | Reply


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: