[ih] Panel lights as a source of computer failure

[Alexander McKenzie]

Steve Crocker suggested that re-engineering the panel lights on the IMP
raised reliability from 98% to 99,98%.  That's not the only fix implemented
to up the reliability, but it was a significant part.  The whole story is
in the article "Seeking High IMP Reliability in the 1970' ARPAnet" by
Walden, McKenzie, and Barker, published in Vol 44, No 2 (April - June 2022)
of IEEE Annals of the History of Computing. The panel light problem was
specific to the second-generation Honeywell-316 IMPs; the original
Honeywell-516 IMPs did not have this problem.

Cheers,
Alex