TypeError: curr.start.toISOString is not a function

KirAsh4

There’s that failure … Notice the contents of curr.start in each iteration:

curr.start is set to:
{ Mon, 01 Jan 1601 02:00:00 GMT tz: undefined }

curr.start is set to:
{ Mon, 01 Jan 1601 01:00:00 GMT tz: undefined }

curr.start is set to:
00) Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna":20160524T203000

That’s where it fails. It processes the previous two just fine, then encounters an invalid time stamp. This is something within the Offce386 calendar. What is that event?

Bangee

its a weekly event.

I’ll Try to delete this event and reload, but thats no the solution… :-(

KirAsh4

It doesn’t appear that the event information is valid to begin with, at least not the way that Office386 is serving the information. It doesn’t make sense to me. It should be an array, not a string. So the next question is, is the data you are receiving valid? And if so, where is that getting mangled that 'curr' ends up with an invalid array.

Unfortunately I don’t have/use Office386’s calendar, so I’m not going to be able to debug this unless you want to share your calendar link with me. or, recreate a blank one with similar events and share that.

KirAsh4

Ok, preliminary digging done. It appears the pattern matching isn’t working correctly for the times that Office386 puts on its calendar, I mean, really 1601 for a year? I don’t think so.

Anyway, I’m going to tag @MichMich on this since he wrote the default calendar and I believe the node-ical stuff as well. I have the output of the ICS, so if you had shared it to get me the link, you can unshare it. I can work with the saved output.

Bangee

ok, sounds not that good as expected. Whats next?

KirAsh4

Basically one of us has to find the time to investigate. The way Office386 creates its dates is really odd. Leave it to Microsoft to break things … as always.

Bangee

ok, my girlfriends calnedar is working. also with office 365 :-) so stupid…

KirAsh4

Does she have the same events as you? Setup the same way? Same time zones configured?

Bangee

she has different events in her calendar, but the all the other things are the same as mine.

MichMich

Do you have a calendar feed which I can test that gives the error?

Bangee

please give me your Mailadress, I’ll share my Calendar with you. :-)

MichMich

Got it. Thanks. Will probably be able to look into this next week. Is that ok?

Bangee

yeah, no problem. today my calendar is running. I’ve made no changes…

strange, I’ve sent you the link to the working one.

Thanks! :hand_splayed_tone1:

Bangee

@MichMich did you find the problem?

MichMich

@Bangee Did not have any time to look into this … I’m sorry! Did I understand it correctly that you managed to solve it with the new calendar feed you sent me?

Bangee

i’ve sent you the working on. i don’t know why, but most of the time, the calendar is working find.
And sometimes there is this bug. I’ve sent you a diff file, with the differences between the files.

Bangee

@MichMich I still have these problems with office365 calender :-(
Did you found a solution for this issue?

MichMich

Did you tryout the develop branch yet? It might include some changes that could help you.

eugenelai

Any update on this? I’m also using office365 calendar and my Calendar randomly stopped working. I’ve tried the node debug.js and just get the below. @MichMich @KirAsh4

pi@rpi3:~/MagicMirror/modules/default/calendar $ sudo node debug.js
Create fetcher ...
Create fetcher done! 
[ { title: 'Max',
    startDate: '1493568000000',
    endDate: '1493654400000',
    fullDayEvent: true,
    firstYear: 1990 },
  { title: 'Someone',
    startDate: '1493740800000',
    endDate: '1493827200000',
    fullDayEvent: true,
    firstYear: 2010 },
  { title: 'Another',
    startDate: '1493913600000',
    endDate: '1494000000000',
    fullDayEvent: true,
    firstYear: 2000 },
  { title: 'Event',
    startDate: '1494086400000',
    endDate: '1494172800000',
    fullDayEvent: true,
    firstYear: 1980 } ]
------------------------------------------------------------

Bangee

nope, still same here… :-(