1998-04-26: Scott all: Implemented logical/physical drivers.. Whew! main.c, mode.c, lcd.* have changed, and I've added text.* and MtxOrb.* driver files... New commandline option "-l driver" selects a driver. 1998-04-21: Scott all: Lots of updates... (patches, etc) Implemented a frame buffer, and custom characters... Added lots of comments, and cleaned a few things... Added a few function to lcd.c 1998-03-22: Scott main.c: Should exit cleanly now, always... It sometimes wouldn't, if killed by init at shutdown... (it would get stuck in a usleep for too long, and then get killed instead of shutting down) main.c: Changed timing method... Sequences now specify their length in TIME_UNITs. A time unit is 1/8th of a second. This should make blinking and other such things easy... main.c: Removed command line options to specify timings... Why would the user really need to change that anyway? main.c: A few mode-feedback things are implemented: blinking, and backlight on/off stuff... mode.c: Xload_screen now sends feedback to toggle blinking for a high load, or turn the backlight off for no load... 1998-03-13: Scott Makefile: Installs man page in /usr/local/man/man1... main.c: Now detects previously running lcdproc. lock.?: New files for detecting another lcdproc... Now detects an already-running process, and gets its pid. So far, it just exits when this happens. 1998-02-28: Scott Makefile: Now handles installation with "make install". main.c: New command line stuff... (in man page) Also, default contrast is 140 now. Looks better from all angles, at least on my machine. :) lcd.1: Man page. New.. :) mode.c: Rearranged memory screen.. Shows Memory and Swap space separately, with two graphs. 1998-02-24: Scott mode.c: Rearranged memory screen. Worked around hbar bios bug. Added swap info to this screen also, so get_mem_info is back to its former state. (call it on an array) mode.c: Fixed xload screen. It was displaying things one space to the left. main.c: Changed command line style from "mode num_times delay_time" to "mode total_time delay_time". Is this better or worse? 1998-02-22: Scott main.c: Now handles SIGINT (Ctrl-C) and SIGTERM (kill) for a clean exit. main.c: Default device now works. (device[] was getting overwritten by argv[0], which would always be "lcd") mode.c: Fixed CPU load meter. Decreased cpu buffer size back to 4. mode.c: Made goodbye_screen() do something... mode.c: Fixed am/pm on clock_screen(). Was previously displaying 12:xx:xxA during the noon hour, instead of 12:xx:xxP. 1998-02-18: Scott mode.c: Fixed crash on memory screen; commented out last two lines of get_mem_info, which accessed a non-existent variable. mode.c: Added "rep" parameter to all screen functions. Specifies how many times that screen has been run. Using 0 will force screen clear/redraw, and the time-based ones also use this to determine if colons should be drawn. main.c: Implemented simple command-line parameters. Example: lcd m 4 .5 /dev/cua1 x 1 2 uses /dev/cua1 for lcd, displaying memory info 4 times at .5 seconds each, and the xload screen once for 2 seconds; then loops. mode.c & lcd.c: created "NOLCD" preprocessor switch. When defined, lcd uses text output (stdout) instead of the com port. mode.c: Changed clock string generation to use "%02d" to achieve that "02:37:06" look. (hint: printf can pad numbers with leading 0's) main.c: Added minimal command-line help.