[calsify] Updated JSCalendar draft version 08

["Robert Stepanek" <[e-mail]>]

All, 
 
I have just uploaded version 08 <https://tools.ietf.org/html/draft-ietf-calext-jscalendar-08> of the JSCalendar RFC draft. This draft is in WG last call and we'd like to send it to the IESG soon. 
 
The draft update includes your feedback on the mailing list and verbal discussion of some mailing list members at CalConnect XLIII:  
 
*Main changes**:* 
 * *Non-mailto URI*: To support all allowed values of iCalendar ORGANIZER and ATTENDEEs, the calendar object's replyTo property now supports non-mailto: URIs. URI templates are not supported anymore. A new Participant.sendTo property using the same value type as replyTo got defined. The email property of a Participant is optional. Participant ids must adhere to the generic id requirements, that is: only contain characters of the "URL and Filename safe" Base64 Alphabet. 
 * *Durations*: the JSEvent duration property got redefined in accordance with iCalendar. The computation of the duration must take time discontinuities in the event's start time-zone into account. The default value got changed from PT0D to PT0S. 
 * *JSON array to JSON set*: to better support partial updates, the type definitions of most multi-valued properties have been changed from JSON arrays to JSON sets. This includes the Participant roles, delegates, memberOf and linkIds properties, as well as the relatedTo property. Recurrence rules are expected to be updated as a single value, and the byXxx properties are still defined as JSON arrays. 
 
*Minor changes:* 
 * *Renamed Participant properties*: the following Participant properties got renamed to avoid ambiguity: participation to attendance, rsvpResponse to participationStatus, rsvpWanted to expectReply. 
 * *Locations*: a location with unspecified time-zone is not required to be in the calendar object's start time-zone. 
 * *Typos*: Fixed a typo in the recurrence rule definition. 
 
*Open points:* 
 * None. A "web" replyTo/sendTo method may be defined, but is not required for submission of the core spec. 
 
Cheers, 
Robert