11:45 p.m.
Le lundi 13 avril 2015 à 21:12 +0200, Patrick Ohly a écrit :
[...]
> #include <boost/function.hpp>
> ^
> compilation terminated.
That just means that some boost dev package is not installed. On Debian,
that is provided by libboost1.55-dev. Install that and compilation
should get further.
libboost1.54-dev in my case (not yet under Debian Jessie), was not
installed.
Configure should already check for libraries, but apparently the boost
check is incomplete.
Can you report back to the list once you have it running?
It compiled successfully with the above library added. But the time
offset remains (not surprisingly 2 hours now, since we are on summer
time).
Problem with the patching? IIRC, it had gone right, but I'd better test
it again.
Best regards,
--
Daniel CLEMENT
12:48 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
On Tue, 2015-04-14 at 08:45 +0200, Daniel CLEMENT wrote:
It compiled successfully with the above library added. But the time
offset remains (not surprisingly 2 hours now, since we are on summer
time).
Problem with the patching? IIRC, it had gone right, but I'd better test
it again.
You can send an event with loglevel=4 to the Funambol server and have a
look at the "generated" item in the html log file. Expand with "+++"
at
the top of the file before searching for it.
--
Best Regards, Patrick Ohly
The content of this message is my personal opinion only and although
I am an employee of Intel, the statements I make here in no way
represent Intel's position on the issue, nor am I authorized to speak
on behalf of Intel on this matter.
2:59 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
Le mardi 14 avril 2015 à 09:48 +0200, Patrick Ohly a écrit :
On Tue, 2015-04-14 at 08:45 +0200, Daniel CLEMENT wrote:
> It compiled successfully with the above library added. But the time
> offset remains (not surprisingly 2 hours now, since we are on summer
> time).
>
> Problem with the patching? IIRC, it had gone right, but I'd better test
> it again.
You can send an event with loglevel=4 to the Funambol server and have a
look at the "generated" item in the html log file. Expand with "+++"
at
the top of the file before searching for it.
Done this. I took care of starting the "right" Syncevolution, since
this
PC now has 2 (one in /usr/bin from repos., and the one compiled
in /usr/local/bin).
The dummy event I created is "test3", below is the contents of the log
file under "generated".
I notice that in the end the TZID is "Europe/Paris" _with quotes_.
Shouldn't the patch have removed them? Maybe I should retry the patch,
paying more attention to its output.
Excerpt of syncevolution-log.html:
--
[2015-04-14 11:46:27.770] Generated:
[2015-04-14 11:46:27.770]
BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Synthesis AG//NONSGML SyncML Engine V3.4.0.47//EN
BEGIN:VTIMEZONE
TZID:Europe/Paris
BEGIN:STANDARD
DTSTART:19671029T030000
RRULE:FREQ=MONTHLY;INTERVAL=12;BYDAY=-1SU
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:19870329T020000
RRULE:FREQ=MONTHLY;INTERVAL=12;BYDAY=-1SU
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
LAST-MODIFIED:20150414T094556Z
DTSTAMP:20150414T094627Z
CREATED:20150414T094556Z
UID:20150414T094551Z-3867-1000-3149-23@e6330d
SEQUENCE:2
CLASS:PUBLIC
TRANSP:OPAQUE
SUMMARY:test3
DESCRIPTION:test3
DTSTART;TZID="Europe/Paris":20150415T150000
DTEND;TZID="Europe/Paris":20150415T153000
END:VEVENT
END:VCALENDAR
--
Daniel CLEMENT
3:04 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
On Tue, 2015-04-14 at 11:59 +0200, Daniel CLEMENT wrote:
I notice that in the end the TZID is "Europe/Paris" _with
quotes_.
Shouldn't the patch have removed them?
Yes.
Maybe I should retry the patch,
paying more attention to its output.
Also double-check that you are using the modified libsynthesis. Make
sure syncevo-dbus-server gets restarted (when using it) or (easier) use
"syncevolution --daemon=no".
--
Best Regards, Patrick Ohly
The content of this message is my personal opinion only and although
I am an employee of Intel, the statements I make here in no way
represent Intel's position on the issue, nor am I authorized to speak
on behalf of Intel on this matter.
4:56 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
Le mardi 14 avril 2015 à 12:04 +0200, Patrick Ohly a écrit :
On Tue, 2015-04-14 at 11:59 +0200, Daniel CLEMENT wrote:
> I notice that in the end the TZID is "Europe/Paris" _with quotes_.
> Shouldn't the patch have removed them?
Yes.
> Maybe I should retry the patch,
> paying more attention to its output.
1) Patching redone. Just to make sure I'm doing nothing wrong: the
patched file
is /syncevolution-1.5/src/synthesis/src/sysync/mimedirprofile.cpp
patched with the syntax
patch mimedirprofile.cpp < syncevo_patch.diff
(the latter being the patch you sent me)
No error.
Also double-check that you are using the modified libsynthesis.
2) How can I ascertain this? I assumed it was enough to start
syncevolution with the full path to the compiled version.
Make
sure syncevo-dbus-server gets restarted (when using it) or (easier) use
"syncevolution --daemon=no".
3) I added that last option to the command line, which read
/usr/local/bin/syncevolution --daemon=no --run loglevel@default=4
funambol
There is still something wrong, because the dummy "test5" event got
synced with time offset, but unlike the previous attempt, I can't see
any trace of it in the logfile. Perhaps point (2) above?
Best regards,
--
Daniel CLEMENT
5:24 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
On Tue, 2015-04-14 at 13:56 +0200, Daniel CLEMENT wrote:
Le mardi 14 avril 2015 à 12:04 +0200, Patrick Ohly a écrit :
> On Tue, 2015-04-14 at 11:59 +0200, Daniel CLEMENT wrote:
> > I notice that in the end the TZID is "Europe/Paris" _with quotes_.
> > Shouldn't the patch have removed them?
>
> Yes.
>
> > Maybe I should retry the patch,
> > paying more attention to its output.
1) Patching redone. Just to make sure I'm doing nothing wrong: the
patched file
is /syncevolution-1.5/src/synthesis/src/sysync/mimedirprofile.cpp
patched with the syntax
patch mimedirprofile.cpp < syncevo_patch.diff
(the latter being the patch you sent me)
No error.
Correct.
> Also double-check that you are using the modified libsynthesis.
2) How can I ascertain this? I assumed it was enough to start
syncevolution with the full path to the compiled version.
You are right, libtool shell wrappers take care of that. However, other
files (like XML fragments) need to be installed unless additional env
variables get set as describe in the main README - I got a bit confused
here myself. Typically I develop with shared libs disabled.
To verify, run
LD_DEBUG=files src/syncevolution --version --daemon=no 2>&1 | grep libsynthesis
That should show your own libsynthesis.
3) I added that last option to the command line, which read
/usr/local/bin/syncevolution --daemon=no --run loglevel@default=4
funambol
There is still something wrong, because the dummy "test5" event got
synced with time offset, but unlike the previous attempt, I can't see
any trace of it in the logfile. Perhaps point (2) above?
Perhaps it wasn't sent because it was already synced before? Did you
recreate or modify it before that sync?
--
Best Regards, Patrick Ohly
The content of this message is my personal opinion only and although
I am an employee of Intel, the statements I make here in no way
represent Intel's position on the issue, nor am I authorized to speak
on behalf of Intel on this matter.
6:08 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
Le mardi 14 avril 2015 à 14:24 +0200, Patrick Ohly a écrit :
[...]
> > Also double-check that you are using the modified libsynthesis.
>
> 2) How can I ascertain this? I assumed it was enough to start
> syncevolution with the full path to the compiled version.
You are right, libtool shell wrappers take care of that. However, other
files (like XML fragments) need to be installed unless additional env
variables get set as describe in the main README - I got a bit confused
here myself. Typically I develop with shared libs disabled.
To verify, run
LD_DEBUG=files src/syncevolution --version --daemon=no 2>&1 | grep libsynthesis
I assume /usr/local/bin/syncevolution (there's no /src/syncevolution) so
I ran:
LD_DEBUG=files /usr/local/bin/syncevolution --daemon=no 2>&1 | grep
libsynthesis
The output:
23521: file=libsynthesis.so.0 [0]; needed
by /usr/lib/libsyncevolution.so.0 [0]
23521: file=libsynthesis.so.0 [0]; generating link map
23521: file=libsqlite3.so.0 [0]; needed
by /usr/lib/libsynthesis.so.0 [0]
23521: file=libexpat.so.1 [0]; needed
by /usr/lib/libsynthesis.so.0 [0]
23521: calling init: /usr/lib/libsynthesis.so.0
23521: calling fini: /usr/lib/libsynthesis.so.0 [0]
... doesn't look good to me: I expected /usr/local/lib/...
^^^^^
How could I force the use of this last path?
Perhaps uninstalling libsynthesis from the repos.?
Perhaps I should even have done this before compiling?
That should show your own libsynthesis.
> 3) I added that last option to the command line, which read
> /usr/local/bin/syncevolution --daemon=no --run loglevel@default=4
> funambol
>
> There is still something wrong, because the dummy "test5" event got
> synced with time offset, but unlike the previous attempt, I can't see
> any trace of it in the logfile. Perhaps point (2) above?
Perhaps it wasn't sent because it was already synced before? Did you
recreate or modify it before that sync?
Yes, I erased the previous dummy event, and recreated this one just
before the last sync (but I did this in one step).
Best regards,
--
Daniel CLEMENT
6:35 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
On 14/04/15 14:08, Daniel CLEMENT wrote:
How could I force the use of this last path?
Like Patrick, I normally debug with shared libs disabled. Unfortunately
library searching has now become very complicated (well beyond my
understanding, particularly with multiarch). However, I have, in the
past, had success using:
LD_LIBRARY_PATH=/usr/local/lib:/usr/lib:/lib
(it needs to be put in front of the syncevolution command in the
commandline).
Try:
LD_LIBRARY_PATH=/usr/local/lib:/usr/lib:/lib \
LD_DEBUG=files \
/usr/local/bin/syncevolution --daemon=no 2>&1 \
| grep libsynthesis
Graham
6:42 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
On Tue, 2015-04-14 at 15:08 +0200, Daniel CLEMENT wrote:
Le mardi 14 avril 2015 à 14:24 +0200, Patrick Ohly a écrit :
> [...]
> > > Also double-check that you are using the modified libsynthesis.
> >
> > 2) How can I ascertain this? I assumed it was enough to start
> > syncevolution with the full path to the compiled version.
>
> You are right, libtool shell wrappers take care of that. However, other
> files (like XML fragments) need to be installed unless additional env
> variables get set as describe in the main README - I got a bit confused
> here myself. Typically I develop with shared libs disabled.
>
> To verify, run
> LD_DEBUG=files src/syncevolution --version --daemon=no 2>&1 | grep
libsynthesis
I assume /usr/local/bin/syncevolution (there's no /src/syncevolution)
There is, in your compile directory. It's a shell script with sets
LD_LIBRARY_PATH such that it uses the uninstalled shared libraries.
I thought you were running without "make install". Once you do that, you
are on your own regarding LD_LIBRARY_PATH. Graham's suggestion should
work.
--
Best Regards, Patrick Ohly
The content of this message is my personal opinion only and although
I am an employee of Intel, the statements I make here in no way
represent Intel's position on the issue, nor am I authorized to speak
on behalf of Intel on this matter.
10:06 a.m.
New subject: [SyncEvolution] Calendar events one hour later than original
Le mardi 14 avril 2015 à 15:42 +0200, Patrick Ohly a écrit :
On Tue, 2015-04-14 at 15:08 +0200, Daniel CLEMENT wrote:
> Le mardi 14 avril 2015 à 14:24 +0200, Patrick Ohly a écrit :
> > [...]
> > > > Also double-check that you are using the modified libsynthesis.
> > >
> > > 2) How can I ascertain this? I assumed it was enough to start
> > > syncevolution with the full path to the compiled version.
> >
> > You are right, libtool shell wrappers take care of that. However, other
> > files (like XML fragments) need to be installed unless additional env
> > variables get set as describe in the main README - I got a bit confused
> > here myself. Typically I develop with shared libs disabled.
> >
> > To verify, run
> > LD_DEBUG=files src/syncevolution --version --daemon=no 2>&1 | grep
libsynthesis
>
> I assume /usr/local/bin/syncevolution (there's no /src/syncevolution)
There is, in your compile directory. It's a shell script with sets
LD_LIBRARY_PATH such that it uses the uninstalled shared libraries.
I thought you were running without "make install". Once you do that, you
are on your own regarding LD_LIBRARY_PATH. Graham's suggestion should
work.
I'm glad to report success at last.
I could check that with the prefix that Graham suggested, the
libsynthesis from /usr/local/ was used.
Then, another test with a dummy calendar item ended in successful sync,
without time offset (so I did not bother to inspect the log).
(I used make install because my 1st idea was to use checkinstall, for
building a package that I would have reused on my main PC. Then, out of
habit, I fell back to make - install.)
Thank you both for your support. If you want me to test further while
I'm at it, just ask.
Best regards,
--
Daniel CLEMENT
2109
days inactive
2109
days old
syncevolution@synevolution.org
9 comments
3 participants
participants (3)
-
Daniel CLEMENT -
Graham Cobb -
Patrick Ohly