[Users] claws-mail displays ics file wrongly

[clawsmail at pasdzierny.de]

hello,

I was looking at that file again and I think I figured:

(1) there is *one* VTIMEZONE only, describing 'central european
time' CET with its various daylight variants including CEST
Its TZID is spezified with Europe/Berlin
(makes sense to me.)

(2) there is *one* VEVENT 
It is giving its DTSTART and DTEND as
  DTSTART:20240827T100000
  DTEND:20240827T101500
nofinal Z, meaning coded as the local timezone
see here: https://icalendar.org/iCalendar-RFC-5545/3-3-5-date-time.html

(3) the timestamp indicating the time of creation of htat VEVENT is
given as
DTSTAMP:20240827T071530Z
final Z means coded as UTC
see here: https://icalendar.org/iCalendar-RFC-5545/3-3-5-date-time.html
requesting the local time was 9:15h => correct :)

(4) I do not understand how exactly the correct DAYLIGHT processing
rule could be derived from that particular ics file. RFC2445 describes
something matching the files structure, but I dont see how that should
work exactly.
I guess the expectation is the interpreting software generally knows
such rules. In this case it should have known that 'CEST' is active and
apply the additional hour.

My botttom line: it is wrong to interprete the VEVENT's times as
UTC because it is not coded with final 'Z', they need to be interpreted
as local time (here "CEST"). 
As long as there was no need to integrate with other data they could
be presented 'as is' plus the given timezone info (here: "CEST")
'Recalculating' the time info as if it where UTC-coded and then
presenting the result explicitly as 'CEST' is definitely wrong.

sincerely, Martin.



On Wed, 28 Aug 2024 13:35:31 +0000
"David Fletcher" <David at megapico.co.uk> wrote:

> >From: clawsmail at pasdzierny.de
> >hello,
> >
> >I am not an expert in ics files, but the attached ics file should
> >refer to 10h CEST (and it does so when imported into google calendar)
> >but I get 12h displayed using claws-mail (version 4.3.0 )(HTML was
> >turned off), see attached screenshot as pdf
> >
> >Because of the difference of 2h is the same as the difference from
> >CEST to UTC my guess is that the timezone is considered falsely or
> >not at all.
> >
> >cheers, Martin.  
> 
> 
> I think there's something strange about this ICS calendar file.
> 
> Germany is 1 hour ahead of United Kingdom. Opening it in the UK it
> shows a meeting at 11am (UK time) in Claws, so that is two hours the
> wrong way. If I open it using the Apple calendar it shows a meeting
> at 10am (UK time), so that is one hour wrong.
> 
> If it's meant to be 10am CEST there's something wrong with the ICS
> file as it should show 9am (UK), not 10 or 11. Central European
> Summer Time is 1 hour ahead of United Kingdom.
> 
> Looking inside the file it has 8 different time-zone offset entries as
> pairs of
> 
> TZOFFSETFROM:
> TZOFFSETTO:
> 
> I would guess that wherever this came from has tried to supply
> pre-calculated times for a range of regions, and this over-supply of
> information is confusing not just Claws but also the Apple calendar
> too.
> 
> Looking inside the ICS file shows a product name "Kommunix
> GmbH//iCal4j 1.0//DE". Searching Google for this takes me to a github
> page: https://github.com/ical4j/ical4j/issues
> 
> Looking quickly (not in-depth) at the issues listed there, it has
> multiple complaints about the timezone information for various regions
> being wrong. Maybe this is the issue, not Claws Mail! Maybe the ICS
> file is theoretically correct, but actually too complicated for some
> calendars (including the Apple one) to interpret.
> 
> Best regards, David.
> _______________________________________________
> Users mailing list
> Users at lists.claws-mail.org
> https://lists.claws-mail.org/cgi-bin/mailman/listinfo/users