2:03 a.m.
Hello!
Daniel reported that auto syncing only runs once since 1.2.99.1. I'm
currently preparing a 1.3.2 update with a fix (and test) for that. See
https://bugs.freedesktop.org/show_bug.cgi?id=56667
Are there any other issues people would like to have fixed in a stable
update?
"Can't synchronise CardDAV with Basic Auth"
https://bugs.freedesktop.org/show_bug.cgi?id=56263
might be one.
Tino, how is it going with updating SyncEvolution in Debian Testing? The
current version in Testing, 1.2.99.1, also has that bug.
--
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.
9:28 a.m.
Hello,
Le lundi 05 novembre 2012 à 11:03 +0100, Patrick Ohly a écrit :
Hello!
[...]
Are there any other issues people would like to have fixed in a stable
update?
Not much of an issue, but I take this opportunity to mention some
cosmetic flaw in the (French) GUI. It's very minor, we almost never see
this window in case of an auto sync.
I'm attaching a screenshot, here are some comments in case it doesn't
reach the list.
The main title somewhat overlaps the "autosync" checkbox and the
"actions" frame. This can be fixed by widening the window, but it
doesn't remember it's last size. This could be fixed by shorter French
translations (Could I help there? -- BTW, the notification messages that
pop up before and after an auto sync are always in English).
While I'm at it, you'll notice that the second "action" button has
centered text in it, whereas the last one is left-justified.
[...]
Forgive me for being so finicky... I just think that Syncevolution is a
great piece of software that deserves a flawless GUI!
Best regards,
--
Daniel CLEMENT
11:51 a.m.
On Tue, 2012-11-06 at 18:28 +0100, Daniel CLEMENT wrote:
Hello,
Le lundi 05 novembre 2012 à 11:03 +0100, Patrick Ohly a écrit :
> Hello!
>
> [...]
>
> Are there any other issues people would like to have fixed in a stable
> update?
Not much of an issue, but I take this opportunity to mention some
cosmetic flaw in the (French) GUI. It's very minor, we almost never see
this window in case of an auto sync.
I'm attaching a screenshot, here are some comments in case it doesn't
reach the list.
The main title somewhat overlaps the "autosync" checkbox and the
"actions" frame. This can be fixed by widening the window, but it
doesn't remember it's last size.
Agreed, that doesn't look right. I would expect it to use a layout that
doesn't allow elements to overlap. However, I am not familiar enough
with GTK to diagnose or fix this.
This could be fixed by shorter French
translations (Could I help there? -- BTW, the notification messages that
pop up before and after an auto sync are always in English).
Yes, help with translating SyncEvolution would be welcome. It used to be
translated for Moblin by Intel, but that effort is mostly dormant
nowadays.
I wonder whether I can get existing translation efforts to look at it.
For example, GNOME has a section for "freedesktop.org" projects
(http://l10n.gnome.org/languages/fr/freedesktop-org/ui/), but
SyncEvolution did not automatically become a part of that and I don't
know where to ask to get it included, nor how this would work in
practice.
While I'm at it, you'll notice that the second
"action" button has
centered text in it, whereas the last one is left-justified.
That looks like an oversight. A "<property
name="justify">center</property>" in ui.xml resp. ui3.xml should
fix
this. Should be doable for 1.3.2 without introducing any risk of
breakage.
Forgive me for being so finicky... I just think that Syncevolution is
a
great piece of software that deserves a flawless GUI!
Thanks for reporting these problems! Every little bit counts.
--
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.
1:15 a.m.
Hello,
Le mardi 06 novembre 2012 à 20:51 +0100, Patrick Ohly a écrit :
[...]
> This could be fixed by shorter French
> translations (Could I help there? -- BTW, the notification messages that
> pop up before and after an auto sync are always in English).
Yes, help with translating SyncEvolution would be welcome. It used to be
translated for Moblin by Intel, but that effort is mostly dormant
nowadays.
FWIW, I had a look at the sources (1.3.1). Found /po/fr.po. Here is an
excerpt:
---
#. TRANSLATORS: This is the title on main view. Placeholder is
#. * the service name. Example: "Google - synced just now"
#: ../src/gtk-ui/sync-ui.c:541
#, c-format
msgid "%s - synced just now"
msgstr "%s - dernière synchronisation il y a quelques secondes"
#: ../src/gtk-ui/sync-ui.c:545
#, c-format
msgid "%s - synced a minute ago"
msgstr "%s - dernière synchronisation il y a moins d'une minute"
---
and so on.
As you can see, the French is considerably longer. In particular,
"synchronisation" is never abridged. I have no simple way to start the
UI in English, but I'd bet it looks better.
Further in the fr.po file they talk about linebreaks allowed. But I
don't know if it can be done here in the main title.
However, I plan to try two modified versions of this fr.po, one with
linebreaks and the other with shortened French labels.
I'll come back and report if I get something conclusive.
[...]
Now for the autosync part. The localization of the notification messages
appears to be addressed in fr.po, e.g.:
---
#. if sync is successfully started and done
#: ../src/syncevo-dbus-server.cpp:6670
#, c-format
msgid "%s sync complete"
msgstr "Synchronisation %s terminée"
---
but, for some reason, the translated strings don't make it into my
notification messages (Linux Mint, Debian, XFCE). I wonder if it's an
XFCE issue: do other users get their notifications translated?
Best regards,
--
Daniel CLEMENT
6:34 a.m.
On Wed, 2012-11-07 at 10:15 +0100, Daniel CLEMENT wrote:
Hello,
Le mardi 06 novembre 2012 à 20:51 +0100, Patrick Ohly a écrit :
> [...]
>
> > This could be fixed by shorter French
> > translations (Could I help there? -- BTW, the notification messages that
> > pop up before and after an auto sync are always in English).
>
> Yes, help with translating SyncEvolution would be welcome. It used to be
> translated for Moblin by Intel, but that effort is mostly dormant
> nowadays.
FWIW, I had a look at the sources (1.3.1). Found /po/fr.po. Here is an
excerpt:
---
#. TRANSLATORS: This is the title on main view. Placeholder is
#. * the service name. Example: "Google - synced just now"
#: ../src/gtk-ui/sync-ui.c:541
#, c-format
msgid "%s - synced just now"
msgstr "%s - dernière synchronisation il y a quelques secondes"
#: ../src/gtk-ui/sync-ui.c:545
#, c-format
msgid "%s - synced a minute ago"
msgstr "%s - dernière synchronisation il y a moins d'une minute"
---
and so on.
As you can see, the French is considerably longer. In particular,
"synchronisation" is never abridged. I have no simple way to start the
UI in English, but I'd bet it looks better.
You can use "LC_ALL=en_US.UTF-8 LANGUAGE=en_US sync-ui" to start the UI
in English.
Further in the fr.po file they talk about linebreaks allowed. But I
don't know if it can be done here in the main title.
I don't know either.
However, I plan to try two modified versions of this fr.po, one with
linebreaks and the other with shortened French labels.
I'll come back and report if I get something conclusive.
>
> [...]
>
Now for the autosync part. The localization of the notification messages
appears to be addressed in fr.po, e.g.:
---
#. if sync is successfully started and done
#: ../src/syncevo-dbus-server.cpp:6670
#, c-format
msgid "%s sync complete"
msgstr "Synchronisation %s terminée"
---
but, for some reason, the translated strings don't make it into my
notification messages (Linux Mint, Debian, XFCE). I wonder if it's an
XFCE issue: do other users get their notifications translated?
No, that seems to be completely broken. I wonder whether no-one has ever
used auto syncing or whether not having translated messages was accepted
as normal.
The explanation is simple: localization was used, but never turned on in
the synevo-dbus-server, the process which creates these notifications. I
have a fix (and a unit test).
Thanks for reporting this.
--
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.
6:24 a.m.
Le vendredi 09 novembre 2012 à 15:34 +0100, Patrick Ohly a écrit :
On Wed, 2012-11-07 at 10:15 +0100, Daniel CLEMENT wrote:
[...]
You can use "LC_ALL=en_US.UTF-8 LANGUAGE=en_US sync-ui" to start the UI
in English.
I did. I found that the overlapping is still there, though much less
obvious that in French, thanks to the shorter English labels.
> Further in the fr.po file they talk about linebreaks allowed. But I
> don't know if it can be done here in the main title.
I don't know either.
> However, I plan to try two modified versions of this fr.po, one with
> linebreaks [...]
I fear I have nothing to report here, as I ran into errors during the
"make" process. Attached error message. I rather would have expected
errors at runtime.
> [...]
Thanks for reporting this.
You're welcome, I wish I could help more actively.
--
Daniel CLEMENT
2:31 a.m.
On Sun, 2012-11-11 at 15:24 +0100, Daniel CLEMENT wrote:
I fear I have nothing to report here, as I ran into errors during
the
"make" process. Attached error message. I rather would have expected
errors at runtime.
In file included from src/gdbusxx/gdbus-cxx-bridge.h:59:0,
from src/gdbusxx/gdbus-cxx-bridge.cpp:20:
src/gdbusxx/gdbus-cxx.h:25:30: erreur fatale: boost/function.hpp : Aucun
fichier ou dossier de ce type
compilation terminée.
boost/function.hpp - file not found?
Do you have the boost development files installed, in particular the one
for boost function? On Debian Testing, that file is installed via
libboost-dev.
The configure script checks for "boost base"; I'm not sure whether that
includes boost/function.hpp.
Hmm, I think the problem is rather the lack of BOOST_CPPFLAGS in the
flags used when compiling gdbusxx.
Does this patch fix the problem? It also adds SYNCEVO_WFLAGS, which was
missing (unrelated, makes no difference unless there are problems in the
code).
diff --git a/src/gdbusxx/gdbusxx.am b/src/gdbusxx/gdbusxx.am
index bd93406..c78d34a 100644
--- a/src/gdbusxx/gdbusxx.am
+++ b/src/gdbusxx/gdbusxx.am
@@ -12,12 +12,13 @@ src_gdbusxx_libgdbussyncevo_la_SOURCES = \
src_gdbusxx_libgdbussyncevo_la_LDFLAGS = $(src_gdbus_version_info)
src_gdbusxx_libgdbussyncevo_la_LIBADD = $(GLIB_LIBS) $(DBUS_LIBS)
-src_gdbusxx_libgdbussyncevo_la_CXXFLAGS = $(GLIB_CFLAGS) $(DBUS_CFLAGS)
+src_gdbusxx_libgdbussyncevo_la_CXXFLAGS = $(GLIB_CFLAGS) $(DBUS_CFLAGS)
$(SYNCEVO_WFLAGS)
+src_gdbusxx_libgdbussyncevo_la_CPPFLAGS = $(BOOST_CPPFLAGS)
MAINTAINERCLEANFILES += Makefile.in
noinst_PROGRAMS += src/gdbusxx/example
src_gdbusxx_example_SOURCES = src/gdbusxx/test/example.cpp
-src_gdbusxx_example_CPPFLAGS = -I$(top_srcdir)/src/gdbusxx/
-src_gdbusxx_example_CXXFLAGS = $(GLIB_CFLAGS) $(DBUS_CFLAGS)
+src_gdbusxx_example_CPPFLAGS = -I$(top_srcdir)/src/gdbusxx/ $(BOOST_CPPFLAGS)
+src_gdbusxx_example_CXXFLAGS = $(GLIB_CFLAGS) $(DBUS_CFLAGS) $(SYNCEVO_WFLAGS)
src_gdbusxx_example_LDADD = src/gdbusxx/libgdbussyncevo.la $(GLIB_LIBS) $(DBUS_LIBS)
--
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.
12:25 a.m.
On Mon, 2012-11-12 at 11:31 +0100, Patrick Ohly wrote:
Hmm, I think the problem is rather the lack of BOOST_CPPFLAGS in the
flags used when compiling gdbusxx.
Does this patch fix the problem? It also adds SYNCEVO_WFLAGS, which was
missing (unrelated, makes no difference unless there are problems in the
code).
[...]
I've added that patch for testing against the 1.3.x branch and will
include it in 1.3.2. However, some confirmation that it really helps
would be useful before claiming to have fixed a problem in the commit
message and NEWS file ;-)
--
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.
10:26 a.m.
Hello,
I've got one good and one bad news:
- with libboost-dev and patched gdbusxx.am, syncevolution 1.3.1 compiled
and installed fine. I think this is the most important thing.
- but I have no luck with my modified fr.po (with line breaks). In
sync-ui I just have no title at all in the main window (replaced by
Cinnamon's "missing" icon).
I'll make further tests, but first I'd like to make sure: do
syncevolution and sync-ui share the same localized files? isn't there
anything in the sync-ui sources that should be modified?
Best regards, Daniel
Le lundi 12 novembre 2012 à 11:31 +0100, Patrick Ohly a écrit :
On Sun, 2012-11-11 at 15:24 +0100, Daniel CLEMENT wrote:
> I fear I have nothing to report here, as I ran into errors during the
> "make" process. Attached error message. I rather would have expected
> errors at runtime.
In file included from src/gdbusxx/gdbus-cxx-bridge.h:59:0,
from src/gdbusxx/gdbus-cxx-bridge.cpp:20:
src/gdbusxx/gdbus-cxx.h:25:30: erreur fatale: boost/function.hpp : Aucun
fichier ou dossier de ce type
compilation terminée.
boost/function.hpp - file not found?
Do you have the boost development files installed, in particular the one
for boost function? On Debian Testing, that file is installed via
libboost-dev.
The configure script checks for "boost base"; I'm not sure whether that
includes boost/function.hpp.
Hmm, I think the problem is rather the lack of BOOST_CPPFLAGS in the
flags used when compiling gdbusxx.
Does this patch fix the problem? It also adds SYNCEVO_WFLAGS, which was
missing (unrelated, makes no difference unless there are problems in the
code).
diff --git a/src/gdbusxx/gdbusxx.am b/src/gdbusxx/gdbusxx.am
index bd93406..c78d34a 100644
--- a/src/gdbusxx/gdbusxx.am
+++ b/src/gdbusxx/gdbusxx.am
@@ -12,12 +12,13 @@ src_gdbusxx_libgdbussyncevo_la_SOURCES = \
src_gdbusxx_libgdbussyncevo_la_LDFLAGS = $(src_gdbus_version_info)
src_gdbusxx_libgdbussyncevo_la_LIBADD = $(GLIB_LIBS) $(DBUS_LIBS)
-src_gdbusxx_libgdbussyncevo_la_CXXFLAGS = $(GLIB_CFLAGS) $(DBUS_CFLAGS)
+src_gdbusxx_libgdbussyncevo_la_CXXFLAGS = $(GLIB_CFLAGS) $(DBUS_CFLAGS)
$(SYNCEVO_WFLAGS)
+src_gdbusxx_libgdbussyncevo_la_CPPFLAGS = $(BOOST_CPPFLAGS)
MAINTAINERCLEANFILES += Makefile.in
noinst_PROGRAMS += src/gdbusxx/example
src_gdbusxx_example_SOURCES = src/gdbusxx/test/example.cpp
-src_gdbusxx_example_CPPFLAGS = -I$(top_srcdir)/src/gdbusxx/
-src_gdbusxx_example_CXXFLAGS = $(GLIB_CFLAGS) $(DBUS_CFLAGS)
+src_gdbusxx_example_CPPFLAGS = -I$(top_srcdir)/src/gdbusxx/ $(BOOST_CPPFLAGS)
+src_gdbusxx_example_CXXFLAGS = $(GLIB_CFLAGS) $(DBUS_CFLAGS) $(SYNCEVO_WFLAGS)
src_gdbusxx_example_LDADD = src/gdbusxx/libgdbussyncevo.la $(GLIB_LIBS) $(DBUS_LIBS)
--
Daniel CLEMENT
10:35 a.m.
On Sun, 2012-11-25 at 19:26 +0100, Daniel CLEMENT wrote:
I've got one good and one bad news:
- with libboost-dev and patched gdbusxx.am, syncevolution 1.3.1 compiled
and installed fine. I think this is the most important thing.
You mention two aspects - which one solved the problem? Installing
libboost-dev or patching gdbusxx.am?
- but I have no luck with my modified fr.po (with line breaks). In
sync-ui I just have no title at all in the main window (replaced by
Cinnamon's "missing" icon).
That sounds like you did "make" without "make install". Could that be
it?
Or perhaps you started the system's sync-ui with a catalog
in /usr/share/locale while your modified version is elsewhere
(like /usr/local)?
I'll make further tests, but first I'd like to make sure: do
syncevolution and sync-ui share the same localized files?
Yes.
isn't there
anything in the sync-ui sources that should be modified?
No, modifying the .mo file is enough to get the texts inside the sync-ui
translated differently.
--
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.
10:53 p.m.
Le dimanche 25 novembre 2012 à 19:35 +0100, Patrick Ohly a écrit :
On Sun, 2012-11-25 at 19:26 +0100, Daniel CLEMENT wrote:
> I've got one good and one bad news:
>
> - with libboost-dev and patched gdbusxx.am, syncevolution 1.3.1 compiled
> and installed fine. I think this is the most important thing.
You mention two aspects - which one solved the problem? Installing
libboost-dev or patching gdbusxx.am?
AFAICT the compilation went further than before with the patched
gdbusxx.am (which I did first), but only ran completely when I installed
libboost-dev.
> - but I have no luck with my modified fr.po (with line breaks). In
> sync-ui I just have no title at all in the main window (replaced by
> Cinnamon's "missing" icon).
That sounds like you did "make" without "make install". Could that
be
it?
No, I did sudo make install at the end.
Or perhaps you started the system's sync-ui with a catalog
in /usr/share/locale while your modified version is elsewhere
(like /usr/local)?
Maybe. After compiling syncevolution, I realized sync-ui was not on this
machine. I was eager to see the result, so rushed and pulled it from the
repos. Maybe it would have been better to compile it also from source.
I'll check the paths carefully to make sure they are consistent.
> I'll make further tests, but first I'd like to make sure: do
> syncevolution and sync-ui share the same localized files?
Yes.
> isn't there
> anything in the sync-ui sources that should be modified?
No, modifying the .mo file is enough to get the texts inside the sync-ui
translated differently.
Best regards,
--
Daniel CLEMENT
2:05 a.m.
On Fri, 2012-11-09 at 15:34 +0100, Patrick Ohly wrote:
On Wed, 2012-11-07 at 10:15 +0100, Daniel CLEMENT wrote:
> Now for the autosync part. The localization of the notification messages
> appears to be addressed in fr.po, e.g.:
> ---
> #. if sync is successfully started and done
> #: ../src/syncevo-dbus-server.cpp:6670
> #, c-format
> msgid "%s sync complete"
> msgstr "Synchronisation %s terminée"
> ---
> but, for some reason, the translated strings don't make it into my
> notification messages (Linux Mint, Debian, XFCE). I wonder if it's an
> XFCE issue: do other users get their notifications translated?
No, that seems to be completely broken. I wonder whether no-one has ever
used auto syncing or whether not having translated messages was accepted
as normal.
The explanation is simple: localization was used, but never turned on in
the synevo-dbus-server, the process which creates these notifications. I
have a fix (and a unit test).
It turned out that this explanation is wrong: localization was turned
on, just not in the place where I was expecting it to be. It is done
inside a helper class, NotificationBackendLibnotify::init(), and not in
main().
That means that I am still unsure why messages were not translated
correctly for you. I was able to reproduce it, but I probably did it
wrong when testing manually (syncevolution.mo not installed, for
example) and the automated test couldn't work (because the path to
syncevolution.mo has to be specified, which wasn't possible).
Daniel, can you run
strace -e trace=file <path>/syncevo-dbus-server | grep
'syncevolution\.mo'
when syncevo-dbus-server is not already running? Does it find the
translated messages?
--
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.
12:02 p.m.
Hello,
Le mercredi 21 novembre 2012 à 11:05 +0100, Patrick Ohly a écrit :
On Fri, 2012-11-09 at 15:34 +0100, Patrick Ohly wrote:
> On Wed, 2012-11-07 at 10:15 +0100, Daniel CLEMENT wrote:
> > Now for the autosync part. The localization of the notification messages
> > appears to be addressed in fr.po, e.g.:
> > ---
> > #. if sync is successfully started and done
> > #: ../src/syncevo-dbus-server.cpp:6670
> > #, c-format
> > msgid "%s sync complete"
> > msgstr "Synchronisation %s terminée"
> > ---
> > but, for some reason, the translated strings don't make it into my
> > notification messages (Linux Mint, Debian, XFCE). I wonder if it's an
> > XFCE issue: do other users get their notifications translated?
>
> No, that seems to be completely broken. I wonder whether no-one has ever
> used auto syncing or whether not having translated messages was accepted
> as normal.
>
> The explanation is simple: localization was used, but never turned on in
> the synevo-dbus-server, the process which creates these notifications. I
> have a fix (and a unit test).
It turned out that this explanation is wrong: localization was turned
on, just not in the place where I was expecting it to be. It is done
inside a helper class, NotificationBackendLibnotify::init(), and not in
main().
That means that I am still unsure why messages were not translated
correctly for you. I was able to reproduce it, but I probably did it
wrong when testing manually (syncevolution.mo not installed, for
example) and the automated test couldn't work (because the path to
syncevolution.mo has to be specified, which wasn't possible).
Daniel, can you run
strace -e trace=file <path>/syncevo-dbus-server | grep
'syncevolution\.mo'
when syncevo-dbus-server is not already running? Does it find the
translated messages?
I'm not familiar with the strace command. It produces a load of output,
but I'm no good at interpreting it. I'm attaching it to this message,
hoping it helps.
Best regards,
--
Daniel CLEMENT
12:12 p.m.
On Wed, 2012-11-21 at 21:02 +0100, Daniel CLEMENT wrote:
Hello,
Le mercredi 21 novembre 2012 à 11:05 +0100, Patrick Ohly a écrit :
> That means that I am still unsure why messages were not translated
> correctly for you. I was able to reproduce it, but I probably did it
> wrong when testing manually (syncevolution.mo not installed, for
> example) and the automated test couldn't work (because the path to
> syncevolution.mo has to be specified, which wasn't possible).
>
> Daniel, can you run
> strace -e trace=file <path>/syncevo-dbus-server | grep
'syncevolution\.mo'
> when syncevo-dbus-server is not already running? Does it find the
> translated messages?
>
I'm not familiar with the strace command. It produces a load of output,
but I'm no good at interpreting it. I'm attaching it to this message,
hoping it helps.
I forget to redirect the strace output into the grep command:
strace -e trace=file <path>/syncevo-dbus-server 2>&1 | grep -e
'syncevolution\.mo' -e INFO -e ERROR
That's make the output a bit smaller.
In this case, syncevo-dbus-server did not try to load any localized
messages. Can you repeat the command and this time wait until an
automatic sync runs? Do you see localized or English notifications?
--
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.
2:07 a.m.
Le mercredi 21 novembre 2012 à 21:12 +0100, Patrick Ohly a écrit :
On Wed, 2012-11-21 at 21:02 +0100, Daniel CLEMENT wrote:
> Hello,
>
> Le mercredi 21 novembre 2012 à 11:05 +0100, Patrick Ohly a écrit :
> > That means that I am still unsure why messages were not translated
> > correctly for you. I was able to reproduce it, but I probably did it
> > wrong when testing manually (syncevolution.mo not installed, for
> > example) and the automated test couldn't work (because the path to
> > syncevolution.mo has to be specified, which wasn't possible).
> >
> > Daniel, can you run
> > strace -e trace=file <path>/syncevo-dbus-server | grep
'syncevolution\.mo'
> > when syncevo-dbus-server is not already running? Does it find the
> > translated messages?
> >
> I'm not familiar with the strace command. It produces a load of output,
> but I'm no good at interpreting it. I'm attaching it to this message,
> hoping it helps.
I forget to redirect the strace output into the grep command:
strace -e trace=file <path>/syncevo-dbus-server 2>&1 | grep -e
'syncevolution\.mo' -e INFO -e ERROR
That's make the output a bit smaller.
In this case, syncevo-dbus-server did not try to load any localized
messages. Can you repeat the command and this time wait until an
automatic sync runs? Do you see localized or English notifications?
Alas no, the notifications are still in English.
Best regards,
--
Daniel CLEMENT
2:33 a.m.
On Thu, 2012-11-22 at 11:07 +0100, Daniel CLEMENT wrote:
Le mercredi 21 novembre 2012 à 21:12 +0100, Patrick Ohly a écrit :
> On Wed, 2012-11-21 at 21:02 +0100, Daniel CLEMENT wrote:
> > Hello,
> >
> > Le mercredi 21 novembre 2012 à 11:05 +0100, Patrick Ohly a écrit :
> > > That means that I am still unsure why messages were not translated
> > > correctly for you. I was able to reproduce it, but I probably did it
> > > wrong when testing manually (syncevolution.mo not installed, for
> > > example) and the automated test couldn't work (because the path to
> > > syncevolution.mo has to be specified, which wasn't possible).
> > >
> > > Daniel, can you run
> > > strace -e trace=file <path>/syncevo-dbus-server | grep
'syncevolution\.mo'
> > > when syncevo-dbus-server is not already running? Does it find the
> > > translated messages?
> > >
> > I'm not familiar with the strace command. It produces a load of output,
> > but I'm no good at interpreting it. I'm attaching it to this message,
> > hoping it helps.
>
> I forget to redirect the strace output into the grep command:
> strace -e trace=file <path>/syncevo-dbus-server 2>&1 | grep -e
'syncevolution\.mo' -e INFO -e ERROR
>
> That's make the output a bit smaller.
>
> In this case, syncevo-dbus-server did not try to load any localized
> messages. Can you repeat the command and this time wait until an
> automatic sync runs? Do you see localized or English notifications?
>
Alas no, the notifications are still in English.
Did you capture the strace output?
--
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.
9:03 a.m.
Well it's still the bulky version of the output, because with the
redirection (which, BTW, I thought you had omitted for a reason) I get:
daniel@m6400 ~ $ strace -e
trace=file /usr/lib/syncevolution/syncevo-dbus-server 2>&1 | grep
'syncevolution\.mo' -e INFO -e ERROR > strace_output.txt
grep: syncevolution\.mo: Aucun fichier ou dossier de ce type
(no such file or directory).
I'm attaching the complete output (without 2>&1).
Best regards - Daniel
Le jeudi 22 novembre 2012 à 11:33 +0100, Patrick Ohly a écrit :
On Thu, 2012-11-22 at 11:07 +0100, Daniel CLEMENT wrote:
> Le mercredi 21 novembre 2012 à 21:12 +0100, Patrick Ohly a écrit :
> > On Wed, 2012-11-21 at 21:02 +0100, Daniel CLEMENT wrote:
> > > Hello,
> > >
> > > Le mercredi 21 novembre 2012 à 11:05 +0100, Patrick Ohly a écrit :
> > > > That means that I am still unsure why messages were not translated
> > > > correctly for you. I was able to reproduce it, but I probably did it
> > > > wrong when testing manually (syncevolution.mo not installed, for
> > > > example) and the automated test couldn't work (because the path
to
> > > > syncevolution.mo has to be specified, which wasn't possible).
> > > >
> > > > Daniel, can you run
> > > > strace -e trace=file <path>/syncevo-dbus-server | grep
'syncevolution\.mo'
> > > > when syncevo-dbus-server is not already running? Does it find the
> > > > translated messages?
> > > >
> > > I'm not familiar with the strace command. It produces a load of
output,
> > > but I'm no good at interpreting it. I'm attaching it to this
message,
> > > hoping it helps.
> >
> > I forget to redirect the strace output into the grep command:
> > strace -e trace=file <path>/syncevo-dbus-server 2>&1 | grep -e
'syncevolution\.mo' -e INFO -e ERROR
> >
> > That's make the output a bit smaller.
> >
> > In this case, syncevo-dbus-server did not try to load any localized
> > messages. Can you repeat the command and this time wait until an
> > automatic sync runs? Do you see localized or English notifications?
> >
> Alas no, the notifications are still in English.
Did you capture the strace output?
--
Daniel CLEMENT
11:12 a.m.
On Fri, 2012-11-23 at 18:03 +0100, Daniel CLEMENT wrote:
Well it's still the bulky version of the output, because with
the
redirection (which, BTW, I thought you had omitted for a reason) I get:
daniel@m6400 ~ $ strace -e
trace=file /usr/lib/syncevolution/syncevo-dbus-server 2>&1 | grep
'syncevolution\.mo' -e INFO -e ERROR > strace_output.txt
grep: syncevolution\.mo: Aucun fichier ou dossier de ce type
Should have been "grep -e 'syncevolution\.mo' -e INFO -e ERROR"
^^
Anyway, there's really no access to syncevolution.mo anywhere. Strange.
I see the access to that happening shortly after opening libnotify. This
is with syncevolution 1.3.1 from syncevolution.org:
[DEBUG syncevo-dbus-server 00:00:01] using libnotify.so.4
open("/usr/share/locale/de_DE/LC_MESSAGES/syncevolution.mo", O_RDONLY) = -1
ENOENT (No such file or directory)
open("/usr/share/locale/de/LC_MESSAGES/syncevolution.mo", O_RDONLY) = 10
open("/usr/share/locale/de_DE/LC_MESSAGES/glib20.mo", O_RDONLY) = -1 ENOENT (No
such file or directory)
open("/usr/share/locale/de/LC_MESSAGES/glib20.mo", O_RDONLY) = 10
Which binaries do you use? Does it happen with 1.3.1 from syncevolution.org?
What is the result of:
env | grep -e LC -e LANG
--
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.
6:37 a.m.
Le vendredi 23 novembre 2012 à 20:12 +0100, Patrick Ohly a écrit :
On Fri, 2012-11-23 at 18:03 +0100, Daniel CLEMENT wrote:
[...]
Should have been "grep -e 'syncevolution\.mo' -e INFO -e
ERROR"
^^
Sorry to waste your time...
Anyway, there's really no access to syncevolution.mo anywhere.
Strange.
[...]
Which binaries do you use?
It's 1.3.1 from Tino Keitel's ppa
Does it happen with 1.3.1 from syncevolution.org?
Aren't they the same? I've put my hands back on the test PC, which will
allow me to test 1.3.1 from syncevolution.org anyhow.
What is the result of:
env | grep -e LC -e LANG
daniel@m6400 ~ $ env | grep -e LC -e LANG
LANG=fr_FR.UTF-8
MDM_LANG=fr_FR.UTF-8
Best regards,
--
Daniel CLEMENT
10:53 a.m.
On Sat, 2012-11-24 at 15:37 +0100, Daniel CLEMENT wrote:
Le vendredi 23 novembre 2012 à 20:12 +0100, Patrick Ohly a écrit :
> On Fri, 2012-11-23 at 18:03 +0100, Daniel CLEMENT wrote:
[...]
> Should have been "grep -e 'syncevolution\.mo' -e INFO -e ERROR"
> ^^
>
Sorry to waste your time...
You didn't :-)
> Anyway, there's really no access to syncevolution.mo
anywhere. Strange.
>
[...]
>
> Which binaries do you use?
It's 1.3.1 from Tino Keitel's ppa
> Does it happen with 1.3.1 from syncevolution.org?
Aren't they the same?
The source is the same, but configuration flags for compilation are
different (for example, syncevolution.org binaries open shared libraries
explicitly via dlopen to work with a variety of library versions) and
packaging is different.
I tried Tino's 1.3.1 binaries and can confirm that localization does not
work in them. That'll allow me to narrow down the differences and find
out why that is. Stay tuned...
--
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.
12:05 a.m.
On Sun, 2012-11-25 at 19:53 +0100, Patrick Ohly wrote:
I tried Tino's 1.3.1 binaries and can confirm that localization
does not
work in them. That'll allow me to narrow down the differences and find
out why that is. Stay tuned...
I understand now what the problem is. The code in the notify backend
which enables localization is incomplete. setlocale() is missing. In the
syncevolution.org binaries that call happens indirectly through the KDE
platform module and therefore localization works.
Now that I understand the root cause better, I can proceed with
releasing the fix in 1.3.2.
--
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.
12:26 a.m.
On Mon, Nov 05, 2012 at 11:03:50 +0100, Patrick Ohly wrote:
Hello!
Daniel reported that auto syncing only runs once since 1.2.99.1. I'm
currently preparing a 1.3.2 update with a fix (and test) for that. See
https://bugs.freedesktop.org/show_bug.cgi?id=56667
Are there any other issues people would like to have fixed in a stable
update?
"Can't synchronise CardDAV with Basic Auth"
https://bugs.freedesktop.org/show_bug.cgi?id=56263
might be one.
Tino, how is it going with updating SyncEvolution in Debian Testing? The
current version in Testing, 1.2.99.1, also has that bug.
I only prepared the packages as of now.
Regarding the autosync issue, how can I check if autosync really works
without modifying any PIM data? Is there some log output somewhere?
Regards,
Tino
1:21 a.m.
On Fri, 2012-11-09 at 09:26 +0100, tino.keitel+syncevolution(a)tikei.de
wrote:
On Mon, Nov 05, 2012 at 11:03:50 +0100, Patrick Ohly wrote:
> Hello!
>
> Daniel reported that auto syncing only runs once since 1.2.99.1. I'm
> currently preparing a 1.3.2 update with a fix (and test) for that. See
> https://bugs.freedesktop.org/show_bug.cgi?id=56667
>
> Are there any other issues people would like to have fixed in a stable
> update?
>
> "Can't synchronise CardDAV with Basic Auth"
> https://bugs.freedesktop.org/show_bug.cgi?id=56263
> might be one.
>
> Tino, how is it going with updating SyncEvolution in Debian Testing? The
> current version in Testing, 1.2.99.1, also has that bug.
I only prepared the packages as of now.
Regarding the autosync issue, how can I check if autosync really works
without modifying any PIM data? Is there some log output somewhere?
You can check with "--print-sessions" that new sessions appear. It also
triggers desktop notifications.
--
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.
2977
days inactive
2998
days old
syncevolution@synevolution.org
22 comments
3 participants
participants (3)
-
Daniel CLEMENT -
Patrick Ohly -
tino.keitel+syncevolution＠tikei.de