mclure@sri-unix (11/14/82)
#N:sri-unix:1000024:000:208 sri-unix!mclure Nov 13 03:46:00 1982 Strange things seem to happen with very large note files. For example, in auto-sequencing mode it appears to wrap around so that one sees notes twice. Other odd things too. Anyone else notice or have a fix?
essick (11/17/82)
#R:sri-unix:1000024:uiucdcs:23300021:004:1058 uiucdcs!essick Nov 17 11:23:00 1982 When you have read the last new note/response in a notesfile, he is presented with the index page again. At that point, unless he wants to read more of that notesfile, he should type "k" or "q" (or "K"/"Q" if you don't want sequencer info for that notesfile updated) and proceed to the next notesfile. If you type another "j"/"J" instead of leaving that notesfile, you will loop through again. This is because the effect of the j/J keys are to from the index page: skip to the first new text in the notesfile. or from a note/response: skip to the next new text in the notesfile. Where "new" is defined as "more recent than the timestamp that was read from your sequencing information file". This defaults to Jan 1, 1970 if the sequencer is not turned on. The net result is that if all you do is sit on the "j" key, you'll loop through that notesfile forever. Occasional use of the "k" key when finished with a notesfile is required. --Ray Essick, U. of Illinois at Urbana-Champaign p.s. This is the same regardless of the size of the notesfile.
notes (11/17/82)
#R:sri-unix:1000024:uiucdcs:23300022:004:975 uiucdcs!notes Nov 17 14:08:00 1982 Boy am I embarassed. The people are SRI are doing things right. The notes code is in error. And after I had publicly babbled about not using the commands correctly. *Blush*. The problem is with seeking and forgetting to case various values. I did all of the later development on a Vax and thus missed a lot of things that are declared "int" and should have been "long". On the 11's, "int" is 16 bits, the Vax has them 32. Some of my address calculations are of the form: long = int + int * int; which on the 11 comes out as: long = short + short*short; the cast to long does not come until too late to save me. After a number of notes in a notesfile, things begin to wrap. *Yech*. I'm working on getting this fixed and should have some results soon. The errors are not likely to manifest themselves unless you have (1) a PDP-11 and (2) very large notesfiles. I'll post the changes as I finish them. --Ray Essick, University of Illinois at Urbana-Champaign