Wierd EPG issues when behind UTC. on VNSI.

Can someone test whether they get this problem to by adjusting their timezone to something in the US. I am in US/Chicago and have really wierd EPG problems which I wrote about at the end of another thread, but thought it was VNSI related.

I am -6 UTC in America/Chicago. My vdr server runs with this timezone as does xbmc (on two machines). When I start xbmc my startup script deletes Epg4.db and TV13.db. XBMC starts and loads the epg correctly over about five minutes. However, after about 20 mins the epg starts to disappear. At one point the initial current program stays current and all programming for about the next seven hours disappears. If I go to Timeline then I can sometimes skip forwards and see programs correctly start again in seven hours time. I assume the initial load has the correct timezone information but when xbmc requests a refresh of the epg it's messing up the timezone request and requesting UTC instead of UTC-Timezone offset.. (Being on summer time I'm not UTC-7).

Yesterday evening, while testing, I loaded the channel epg which started six hours later.

I only have VDR to test this with (looked at tvheadend last night - but didn't get very far), but really would like someone who has this working (probably with a UTC or UTC+1 timezone) to test it by changing their system tz to here?

Rob

does this problem also occur if you ignore database for epg?

Yes. No change, (except the bar across the top when loading the epg doesn't happen).

Some more debug... I changed my backend to UTC instead of US/Chicago, and the same problem persisted. I changed the backend back to Chicago, and the Frontend to London - the epg was in, and stayed in the correct place, correct order etc...

I don't know at this stage if it's an XBMC or Plugin error. I played a little with tvheadend, but not long enough to get an EPG associated with a channel, so couldn't test it, and I can't get Myth to login to myth, so have given up comparative testing.

please pastebin a full debug log.