Hi people,
I have a bug that took me a while to track down. This bug is that
opie's datebook hangs for ever on a datebook.xml created by multisync
from a evolution file. I believe that the problem is that some
entries from Evolution are not complete, multisync lets this pass
(Garbage In, Garbage Out), and datebook fails to realize the entry is
incomplete.
The entry in evolution's ~/evolution/local/Calendar/calendar.ics looks like:
BEGIN:VEVENT
UID:[EMAIL PROTECTED]
DTSTAMP:20021003T145245Z
DTSTART;TZID=/softwarestudio.org/Olson_20011030_5/Europe/Paris:
20021007T153000
DTEND;TZID=/softwarestudio.org/Olson_20011030_5/Europe/Paris:
20021007T183000
TRANSP:OPAQUE
SEQUENCE:9
SUMMARY:Compil
LOCATION:EPITA
CLASS:PUBLIC
RRULE;X-EVOLUTION-ENDDATE=20021209T143000Z:FREQ=WEEKLY;COUNT=10;
INTERVAL=1
EXDATE;VALUE=DATE:20021111
END:VEVENT
Compared to other repeting events (that do work), there is no BYDAY field.
Multisync transforms this entry into
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE DATEBOOK><DATEBOOK>
<events>
<event uid="173030560" description="Compil" note="Compil" location="EPITA"
start="1034008200" end="1034017200" created="1033649565" rtype="Weekly" rfreq="1"
rhasenddate="0" />
</events>
</DATEBOOK>
note that there is no "rweekdays" in it. When I add one, taking
inspiration from another entry that does work:
<event uid="1869308419" description="Piscine Prenatale" note="Piscine Prenatale"
start="1062693000" end="1062707400" created="1063777943" rtype="Weekly" rfreq="1"
rhasenddate="0" rweekdays="8" />
then magically datebook works properly.
I suspect that evolution is now "fixed": I no longer have entries like
this in more recent dates, so maybe it's done. Nevertheless, I think
that evolution should upgrade older entries to have complete fields.
Multisync should probably be more strict and at least emit a few
warnings for such incomplete fields.
Finally, it is abnormal from opie's datebook to hang indefinitely on
the small datebook.xml I posted above, so it should be fixed too.
Thanks to you all. As a disclaimer I wish to emphasize that I may
have incorrectly suspect some package to have performed something
invalid (e.g., evolution may be fully right to have such records).
But I'm only a user, so that's only a wild guess.
Akim
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Multisync-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/multisync-users
