paul@321 | 1 | iCalendar Serialisation
|
paul@321 | 2 | -----------------------
|
paul@321 | 3 |
|
paul@321 | 4 | The vContent serialisation should probably be used instead of the existing
|
paul@321 | 5 | code.
|
paul@321 | 6 |
|
paul@290 | 7 | Navigation Controls
|
paul@290 | 8 | -------------------
|
paul@290 | 9 |
|
paul@290 | 10 | The "New event" link should probably not be present when only remote events
|
paul@290 | 11 | are being aggregated by a calendar.
|
paul@290 | 12 |
|
paul@218 | 13 | Points in Time
|
paul@218 | 14 | --------------
|
paul@218 | 15 |
|
paul@298 | 16 | Make events with identical start and end times appear in the day view,
|
paul@298 | 17 | potentially using the approach given below.
|
paul@298 | 18 |
|
paul@218 | 19 | Events which have identical start and end times might be represented by
|
paul@218 | 20 | building a calendar scale that distinguishes between times acting as start
|
paul@290 | 21 | times and times acting as end times. (The iCalendar specification appears to
|
paul@290 | 22 | state that events without end dates/times are actually points in time, but
|
paul@290 | 23 | this potentially conflicts with the expectation that merely specifying a start
|
paul@290 | 24 | date or time produces an event with an undefined end point or a "common sense"
|
paul@290 | 25 | end point.)
|
paul@218 | 26 |
|
paul@218 | 27 | Consider making dates convertible to timespans of the form (start of day,
|
paul@218 | 28 | start of next day).
|
paul@218 | 29 |
|
paul@211 | 30 | Localised Keywords
|
paul@211 | 31 | ------------------
|
paul@211 | 32 |
|
paul@211 | 33 | It should be possible to define events using localised equivalents of "start",
|
paul@211 | 34 | "end", "summary" and so on. To achieve this, the page language would be found
|
paul@211 | 35 | and regular expressions built to use the localised keywords, falling back on
|
paul@211 | 36 | the English keywords, would then search for event details.
|
paul@211 | 37 |
|
paul@211 | 38 | Recurring Events
|
paul@211 | 39 | ----------------
|
paul@211 | 40 |
|
paul@232 | 41 | Recurring event information from iCalendar sources should be considered in
|
paul@232 | 42 | order to avoid showing incomplete or incorrect event datetimes. Ultimately,
|
paul@232 | 43 | such information may need to be parsed and incorporated into the general event
|
paul@232 | 44 | recurrence processing.
|
paul@232 | 45 |
|
paul@211 | 46 | Having events recur at certain intervals would potentially involve the
|
paul@211 | 47 | expansion of events to produce multiple instances within a specified period of
|
paul@211 | 48 | interest, and such expansion could occur after an event's details have been
|
paul@211 | 49 | read. Care would need to be taken in cases where no limits are placed on a
|
paul@211 | 50 | calendar: the expanded instances should not be allowed to recede into the past
|
paul@211 | 51 | and future indefinitely; where no other events exist to provide implicit
|
paul@211 | 52 | limits, some other default limits might be required to let the expansion
|
paul@211 | 53 | occur.
|
paul@211 | 54 |
|
paul@211 | 55 | The description of recurring events could be based on the iCalendar
|
paul@211 | 56 | specification, although simpler schemes could be preferable. Recurring event
|
paul@211 | 57 | descriptions might start with "every" and then provide a time period ("day",
|
paul@211 | 58 | "week", "month", "year") for offsets from a specified date or time, perhaps
|
paul@211 | 59 | using qualifiers ("first", "second", "other", and so on), or instead provide a
|
paul@211 | 60 | more complete description using additional qualifiers that may override any
|
paul@211 | 61 | specified date or time for instances other than the primary occurrence. For
|
paul@211 | 62 | example, "every second Wednesday of every other month".
|
paul@211 | 63 |
|
paul@227 | 64 | Possible grammar:
|
paul@227 | 65 |
|
paul@227 | 66 | <recurrence> [ of <recurrence> ]...
|
paul@227 | 67 | [ from <datetime> ]
|
paul@227 | 68 | [ until <datetime> ]
|
paul@227 | 69 |
|
paul@227 | 70 | recurrence = every [ <qualifier> ] <interval>
|
paul@227 | 71 | interval = second | minute | hour | day | <weekday> | week | month | year
|
paul@227 | 72 |
|
paul@227 | 73 | The resolution of each successive <recurrence> must be lower than those it
|
paul@227 | 74 | follows. Thus, "every second day of every second week..." is valid whereas
|
paul@227 | 75 | "every second week of every second day..." is not.
|
paul@227 | 76 |
|
paul@204 | 77 | Map Views
|
paul@204 | 78 | ---------
|
paul@204 | 79 |
|
paul@218 | 80 | Other projections might be supported. This would be necessary for various
|
paul@218 | 81 | retrieved map images.
|
paul@218 | 82 |
|
paul@204 | 83 | Dynamic images obtained from other sites or generated locally might provide some
|
paul@204 | 84 | enhancements to the map view. For example, a weather/radar image might show the
|
paul@204 | 85 | cloud or rain forecast either for the current situation or, if forecasts are
|
paul@204 | 86 | available, for the times of events shown.
|
paul@204 | 87 |
|
paul@204 | 88 | Consider having day numbers down the side of a map view with highlighted days
|
paul@204 | 89 | indicating days having events, and with pop-up elements shown upon hovering over
|
paul@204 | 90 | each highlighted day.
|
paul@204 | 91 |
|
paul@204 | 92 | To Do Items
|
paul@204 | 93 | -----------
|
paul@204 | 94 |
|
paul@204 | 95 | Consider adding support for "to do" items. These might have time-related details
|
paul@204 | 96 | such as deadlines, but are more likely to have relationships with other items,
|
paul@204 | 97 | potentially forming a hierarchy of items.
|
paul@204 | 98 |
|
paul@204 | 99 | Event Section Parser
|
paul@204 | 100 | --------------------
|
paul@204 | 101 |
|
paul@336 | 102 | Events in page sections/regions could be presented using more sophisticated
|
paul@336 | 103 | methods and potentially be editable. To support direct editing, the parser
|
paul@336 | 104 | would provide a hidden form field indicating the location of the section in
|
paul@336 | 105 | the Wiki text, and the new event action would be enhanced to read existing
|
paul@336 | 106 | events from the indicated page region, populating the form fields with the
|
paul@336 | 107 | data found in the page.
|
paul@242 | 108 |
|
paul@242 | 109 | UID Properties
|
paul@242 | 110 | --------------
|
paul@242 | 111 |
|
paul@242 | 112 | Especially in the case of aggregation from multiple sources, the only reliable
|
paul@242 | 113 | way of avoiding repetition of the same events described in different places is
|
paul@242 | 114 | for authors to include a UID property identifying each event, using the same
|
paul@242 | 115 | value regardless of where the event is being published.
|
paul@265 | 116 |
|
paul@292 | 117 | Formatting in iCalendar Output
|
paul@292 | 118 | ------------------------------
|
paul@292 | 119 |
|
paul@292 | 120 | If there is a reasonably standard way of incorporating Wiki text in iCalendar
|
paul@292 | 121 | output alongside plain text, this would enable events aggregated from Wiki
|
paul@292 | 122 | sources to use Wiki text to describe things like the location and topics of an
|
paul@292 | 123 | event with links and other formatting that could then be reproduced in the
|
paul@292 | 124 | aggregating Wiki.
|
paul@292 | 125 |
|
paul@265 | 126 | Remote Source Timeouts
|
paul@265 | 127 | ----------------------
|
paul@265 | 128 |
|
paul@265 | 129 | Sometimes, network problems can cause delays in accessing remote sources. The
|
paul@265 | 130 | library should support either a timeout mechanism or asynchronous retrieval of
|
paul@265 | 131 | remote source data.
|