Hi Michal,
On 28/6/26 23:02, Michal Schorm wrote:
Hello,
I am the maintainer of the package 'mariadb-connector-c'.
I am sorry for your issue, thanks for letting me know.
I analyzed your issue based on your description. It seems to be an
unfortunate regression in the upstream code that causes the temporal
data type to be stored with zero length, resulting in the symptoms you
are observing.
Thanks for looking into this.
You say symptoms but it was a real shock for me. It took my MythTV down hard.
BTW, I guess MythTV uses the c API, it logged a zillion repeats of messages
like this:
2026-06-28T17:52:49+10:00 mythbackend[26647]: 2026-06-28 17:52:49.198163 E
[26647/26747] Scheduler mythdb.cpp:227:DBError DB Error (addHistory):
2026-06-28T17:52:49+10:00 mythbackend[26647]: Query was:
2026-06-28T17:52:49+10:00 mythbackend[26647]: REPLACE INTO oldrecorded
(chanid,starttime,endtime,title,subtitle,description,season,episode,category,seriesid,programid,inetref,findid,recordid,station,rectype,recstatus,duplicate,reactivate,generic,future)
VALUES(?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?);
2026-06-28T17:52:49+10:00 mythbackend[26647]: Driver error was [2/1048]:
2026-06-28T17:52:49+10:00 mythbackend[26647]: QMYSQL: Unable to execute
statement
2026-06-28T17:52:49+10:00 mythbackend[26647]: Database error was:
2026-06-28T17:52:49+10:00 mythbackend[26647]: Column 'starttime' cannot be null
constantly until the package was downgraded, then it stopped.
I will need MariaDB upstream to verify this and create a proper fix.
Until then, I attempted to create a downstream patch:
https://src.fedoraproject.org/rpms/mariadb-connector-c/pull-request/27
Could you please test the PR scratch build and check whether it solves
the issue for you?
Rawhide scratch build:
https://koji.fedoraproject.org/koji/taskinfo?taskID=147147264
F44 scratch build: https://koji.fedoraproject.org/koji/taskinfo?taskID=147147459
I understand how it will help, but unfortunately I am not in a position to do
this.
I am dealing with a MythTV issue and do not want to get side tracked now.
I will avoid updating these packages for a while though. I wonder if this
version should be pulled until fixed.
Eyal
Michal
--
Michal Schorm
Senior Software Engineer
Databases Team
Red Hat
--
On Sun, Jun 28, 2026 at 1:35 PM <[email protected]> wrote:
This version picked by my server today. fc44 on x86_64.
I do not know what is responsible but downgrading to 3.4.8 solved the problem.
Maybe a clash between rpmfusion and fedora?
It looks like it may me a major problem with handling date/time. All "Previously
Recorded" show a date in 2106...
MythWeb, which directly accesses the db, works just fine. myth:6544 has the
problems, as does mythfrontend, and likely mythbackend (no recording made).
Status shows "no upcoming".
All schedules show start/end of 00:00:00. No "Program Guide" data showing. No
"Upcoming" (natch), and more.
This update was in stable for two days now but I do not see any mention, maybe
it is something else on my server?
--
Eyal at Home ([email protected])
--
Eyal at Home ([email protected])
--
_______________________________________________
users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
Fedora Code of Conduct:
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/[email protected]
Do not reply to spam, report it:
https://forge.fedoraproject.org/infra/tickets/issues/new
