CALDAV timeshift when using Google Calendar with the Desktop Browser

I have configured two events. First one using the desktop browser from 14:30 to 15:30, the item unfortunately receives 13:30 as start time:
CALDAV_Start1 changed from UNDEF to 2020-11-19T13:30:00.000+0100
CALDAV_End1 changed from UNDEF to 2020-11-19T14:30:00.000+0100
(note that on android calendar app it is displayed correctly with 14:30 as starting time)

But when I use the calendar app on my android phone, I configured a second event from 16:15 to 17:15, the correct item values are received:
CALDAV_Start2 changed from UNDEF to 2020-11-19T16:15:00.000+0100
CALDAV_End2 changed from UNDEF to 2020-11-19T17:15:00.000+0100

I am currently on the master branch 2.5.10-1.

First I thought that it might be the caldavio:timeZone=Europe/Vienna setting, but changing this did not help.
Bug?

thx!

The browser isn’t likely to matter. What might matter is the UI you’re using.
Seems like the UIs being used are treating the date value inconsistently.

Any problems with the CALDAV binding can’t be fixed, so you’re going to be out of luck unless this can be shown to be a problem with the OH UI you’re using.

It’s not the UI since already the Item has the wrong value. That’s why I included the logging lines …

In the meantime I found the issue:
DTSTART:20201119T133000Z

The .ics data (when entered through desktop browser) is coming without timezone, where the correct event (entered through mobile phone) is coming with timezone:
DTSTART;TZID=Europe/Vienna:20201119T161500

For this I configured the timezone in caldavio.cfg, but it does not take any effect:
caldavio:timeZone=UTC

I also tried this:
caldavio:family:timeZone=UTC

As a side note, might be relevant, when I configure the caldavio:charset=UTF-8 the caldavio.cfg cannot be parsed and errors.

What do you mean with “can’t be fixed”? Is the maintainer gone?
The code should not be too complicated parsing an .ics file …

I would be willing to have a look at this. Where can I find the sources for this binding?

What do you mean with “can’t be fixed”? Is the maintainer gone?
The code should not be too complicated parsing an .ics file …

It can’t be fixed because it’s an OH1 binding, and none of those bindings can be updated. It’s a dead repository.

As far as parsing, I don’t believe the binding does any. It relies on the ical4j package to get data from CALDAV.