AaronJowitt / Rivendell

RIvendell Automation Issue/Req Tracking
0 stars 0 forks source link

Block TS - last event of the day #12

Open AaronJowitt opened 8 years ago

AaronJowitt commented 8 years ago

How do we deal with last hour of the day with block TS. Only way it seems we can enable TS, is to put an externally scheduled HT marker at 23:59:59, but this leaves us 1 second short of midnight. AN internally scheduled HTM can be scheduled at 23:59.59.9, but besides being messy, this leaves us 0.1s short of midnight. How do we TS to midnight, with last items transition (segue start) point occurring at exactly 0:00:00?

Note this would be acceptable, but a short/mid term solution.

For example, on one station we currently play a TS block from 23:00:05 to 00:00:05. Last event of the day is a ~18min syndicated program cart. Its transition (segue start point) hits a HTM at 00:00:05, when the first audio event of the new day is a news theme and we cross to live satellite news. As it currently stands, we will have to temporarily dump our 12am news when transitioning to RD, as we have no way to 'neatly' time the last segment of the previous program out to the news theme at 00:00:05.

Ideally long term, we should hopefully see RD capability improved to be able to TS to a point beyond midnight - and we can put the news back in :-)

ElvishArtisan commented 8 years ago

| Note this would be acceptable, but a short/mid term solution.

But you have this tagged as being deployment critical. Which is it?

AaronJowitt commented 8 years ago

Good question. With present automation we have zero gap around midnight, and its desirable we don't introduce one with RD. We won't be using the internal scheduler (other than to import all external music/traffic), but we could temporarily use the internal scheduler to schedule a target time marker at 23:59:59.9, leaving us a ~100ms gap. Yes - we could live with that for a month or so. I've changed the label to wish list med tag. Perhaps allowing the last time marker to have a 00:00:00 target value may be a quick fix?