Comments on man/calendar.texi.

[Matt Hodges]

I've read through man/calendar.texi and have some comments.

(1) The documentation of calendar-week-start-day doesn't mention that
    if this variable is changed, redraw-calendar needs calling for the
    effect to be visible. Also, until the calendar is redrawn, the
    movement commands (e.g., calendar-forward-day) don't work
    correctly. This situation is confusing and should at least be
    mentioned in the variable documentation; better still if
    customizing the variable redraws the calendar automatically.

(2) The resizing of the calendar window that can occur with, for
    example, calendar-forward-month can be distracting, so a
    customizable minimum window height may be useful.

(3) In Calendar mode, SPC is bound to scroll-other-window, so binding
    DEL to scroll-other-window-down seems natural (cf. view-mode).

(4) The error message "Months before February, 1 AD are not available"
    looks a bit strange if you do, for example, C-u calendar RET 1 RET
    February RET then M-v (scroll-calendar-right-three-months). I
    guess that the message refers specifically to using C-u M-x
    calendar, but there are other routes to generate-calendar, and
    maybe something like "Months before January, 1 AD cannot be
    displayed" would be more informative.

(5) The earliest Mayan long count date appears to be 7.17.18.13.3,
    rather than 7.17.18.13.1. (This needs to be checked.)

(6) Mention timeclock-change, and fix the documentation for the
    default behaviour of timeclock-ask-before-exiting.

See attached patch for suggested changes. Thanks,

Matt

calendar.diff
Description: Calendar patch.