[horde] Problem kronolith
R.J.Baart at Prompt.nl
R.J.Baart at Prompt.nl
Wed Feb 25 13:07:50 UTC 2026
Amazing, I really think I have a huge influence.
The bug in Thunderbird seems to have been fixed this morning in the
latest version of Thunderbird. It concerns Thunderbird Desktop Version
148.0, released February 24, 2026. I haven't checked it yet, but the
release notes say “fixed: iCal imports misread unknown time zones as
GMT, creating events at wrong times.”
That's really fast.
Op 2026-02-25 om 11:02 schreef Ruud Baart:
> Thanks, I made a bug report in Bugzilla. Hope they find a solution.
>
> My (ugly) temporary work around for the moment is to make a progresql
> cronjob that looks for null values in the kronolith_events table and
> fill these with our timezone (Europe/Paris) and adjust the
> modification time. With a sync of the calendar it should be fine for
> the users.
>
>
> Op 2026-02-25 om 10:54 schreef Ralf Lang:
>> Hi Ruud,
>>
>> yes, I read it that way. Now telling Thunderbird to change is beyond
>> me unfortunately. It's a wholly different beast I fear.
>>
>> Ruud Baart <r.j.baart at prompt.nl> schrieb am Di., 24. Feb. 2026, 15:40:
>>
>> Thank you for this explanation.
>>
>> Are you referring to this in the RFC 5545:
>>
>> "The use of local time in a DATE-TIME or TIME value without the
>> "TZID" property parameter is to be interpreted as floating time,
>> regardless of the existence of "VTIMEZONE" calendar components in
>> the iCalendar object."
>>
>> This means: no timezone -> local time. And what local time is that
>> depends on where you are. Correct?
>>
>> It also means the Thunderbird calendar cache is faulty. It should
>> not store "UTC" if timezone is empty. Correct?
>>
>>
>>
>> Op 2026-02-24 om 15:22 schreef Ralf Lang:
>>> Hi Ruud,
>>>
>>> it's a good idea to increase data quality but it also introduces
>>> some new problems.
>>> |VEVENT| components without explicit |VTIMEZONE| entries are
>>> valid in iCalendar (RFC 5545) and CalDAV, provided they follow
>>> specific formatting rules for their time values.
>>> We have to accept events, ics files, invitations etc from other
>>> software products or from legacy data imports.
>>>
>>> However we can define that we internally handle them with
>>> timezone and always create/export them with timezone.
>>> Which leads to a whole other category of fun whenever one
>>> software product uses a timezone another doesn't know.
>>>
>>> We will probably need to work on how exactly we achieve this
>>> timezone compatibility the right way.
>>>
>>>
>>> On Tue, Feb 24, 2026 at 3:13 PM Ruud Baart <r.j.baart at prompt.nl>
>>> wrote:
>>>
>>> Hi,
>>>
>>> approximately 30,000 appointments are stored in the Horde
>>> database,
>>> roughly half of which do not have a time zone stored. That
>>> works fine in
>>> itself, but lately it has been causing problems in Thunderbird.
>>>
>>> horde6=# select count(*) from kronolith_events where
>>> event_timezone is null;
>>> count
>>> -------
>>> 13677
>>> (1 row)
>>>
>>> horde6=# select count(*) from kronolith_events where
>>> event_timezone is not null;
>>> count
>>> -------
>>> 17499
>>> (1 row)
>>>
>>> If I have analyzed it correctly, this is because
>>> Thunderbird's calendar
>>> converts the null value to UTC in its cache (sqlite
>>> databases). That is
>>> not the correct value. So registering an event works fine,
>>> but when you
>>> reopen Thunderbird, the display is incorrect.
>>>
>>> I would like to see Horde make a time zone mandatory when
>>> storing
>>> events. If the time zone is missing, it should take the time
>>> zone from
>>> the preferences, and if that is not possible, it should take
>>> the system
>>> time zone. I am not sure if my wish is already achievable
>>> through the
>>> configuration of Horde.
>>>
More information about the horde
mailing list