That's a very painful story, and a good reminder that when you wire up something 'unexpected' (such as monkey brains) to a computer you should probably not rely on simply taping off a switch or a cardboard sign.
The poor service guy is probably still having nightmares over this, if he's still alive.
What bugs me though is how the patterns written to that disk managed to get sent out to the DACs if the machine was being serviced and running diagnostics, after all, simply writing to the drive would not automatically cause the DACs to be written to.
edit: got it, the DACs where connected right over the top of the drive controller.
My first reaction to this was to laugh - it's really rather funny/unbelievable, when you're looking back on this incident with all the years for perspective. But then I attempted to put myself in the DFS engineer's shoes and didn't like what I saw. Poor soul. This is one maintenance incident he's unlikely to forget.
"At this writing, the Jargon File claims the incident actually happened, at Toronto in 1979 or 1980, and that the sysadmin on duty was actually interviewed. The account doesn't provide enough details to track down an independent account, however.
Current University of Toronto sysadmins have expressed skepticism. For one thing, in almost all versions of the story, including the ostensibly documented one in the Jargon File, the computer is a VAX; at the time a VAX would have been a very unusual platform for this kind of data acquisition (they used PDP-11s). The Toronto zoology department has never been licensed to work with primates; the only section of the university that could have done experiments of this nature was the School of Medicine. Investigation continues."
Actually, the original story does say it was Medicine buying the Vax and doing the experiment, with Zoo helping. The VAX 11-780 was the hot machine of 1979/80 ("a 1 MIPS beast!", I heard).
I've seen Laura Creighton's name on Python lists - I believe she's on LinkedIn. The author's "Statistically Invalid Sampling of My Life" (http://edp.org/index.htm) is also pretty amusing - don't think he'd need to embellish that incident.
The poor service guy is probably still having nightmares over this, if he's still alive.
What bugs me though is how the patterns written to that disk managed to get sent out to the DACs if the machine was being serviced and running diagnostics, after all, simply writing to the drive would not automatically cause the DACs to be written to.
edit: got it, the DACs where connected right over the top of the drive controller.