chain printer speed?

On 6/13/2006 at 5:07 PM Joe R. wrote: Chuck, before his first cuppa said: worked. We're actually saying the same thing--it's just that I forgot to say "matches the order of characters on the chain (or the non-matching characters are spaces), you've printed the line in one swell foop, you're all done and the paper advances. Otherwise, the unprinted positions wait for their respective characters to arrive..." Sometimes, I don't do so well translating the pictures in my head to words. :( One could always hear a job banner being printed because of the peculiar pattern of repeated characters printed across the page perf used as a visible divider (if you looked at a stack of job output, you could pick out individual jobs because normal output didn't print over the perfs. Made life easy for the I/O clerks.) Similarly, you always knew when a CE was working on the printer because of the distinctive sound of the diagnostic test patterns. Type slugs wore at different rates, depending on the application. Since we were always printing lots of deadstart dumps (a Cyber 74 with 4M words of ECS could easily occupy most of a box of paper), the "zeroes" got fuzzy far faster than any other character. I never understood how the band printer uses dealt with this (other than replacing the band regularly). Maybe it's a trick of the eye, but the old drum printers often exhibited a lot of vertical misalignment of characters, making for wavy lines and very annoying reading. Curiously, the horizontal misalignment of characters in chain, train and band printers didn't seem to matter nearly as much. I remember that the Univac 1108 hardware manuals were drum printed and then reproduced and were darned near illegible after a couple of hours of miserable reading. By comparison, 1403 output was pristine. Cheers, Chuck