laser-lovers@uw-beaver (06/08/85)
From: stanonik@nprdc (Ron Stanonik) A recent message to this list, emphasizing that there are 72.27 (not 72) points per inch, prompted me to take another look at a problem with our symbolics laser printer. Without some kludging, troff output wasn't correctly positioned on a page; ie, output for the first page was correctly positioned, but output for the second drifted slightly higher, and the drift increased with each succeeding page. The kludges took one of two forms, either specifying a slightly larger page length (.pl 11.042i), or issuing cut marks which were converted (by catlgp) to page commands. The problem seemed to be that catlgp (the filter converting cat codes to lgp commands) and troff didn't agree on the length of an 11 inch page, expressed in cat units. catlgp assumed 72.27 points per inch and 2 vertical cat units per point, while troff seemed to assume 144 vertical cat units per inch. Changing catlgp (actually PTperIn in units.h from 72.27 to 72.), has allowed us to do away with wiring wierd pagelengths into macro packages (making them unusable on other printers) and to also do away with cut marks (about which other printers complained). I'm not sure who gets the blame here, troff for implicitly using 72 points per inch, or catlgp for using points as an intermediate unit rather simply taking troff's word; ie, 144 vertical cat units per inch. Ron Stanonik stanonik@nprdc