1:52 p.m.
On Mi, 2011-07-27 at 10:30 +0200, Chris Kühl wrote:
On Wed, Jul 27, 2011 at 9:07 AM, Patrick Ohly
<patrick.ohly(a)intel.com> wrote:
> On Mi, 2011-07-27 at 01:45 +0200, Chris Kühl wrote:
>> I've tried my dbus-server-reorg branch with this change as well as
>> your dbus-server-reorg-pohly-libsynccommon branch but I still get the
>> same results as before.
>
> Just to be sure, you mean the "no backends available in static
> compilation" problem?
>
> The last commit fixed it for me. Anyway, the little benefit of compiling
> the source files less often is not worth the hassle. I'll revert to
> including the core source files in all executables.
>
This is the error I get for the TestConnection.testCredentialsRight for example,
DBusException: org.syncevolution.InvalidCall: invalid value 'file' for
property 'backend': 'not one of the valid values (virtual, calendar =
events, addressbook = contacts, todo = tasks, memo = memos = notes)'
I forgot that we were talking about the syncevo-dbus-server. My patch
was only for the syncevolution binary. Because of its ugliness I must
have shied away from applying the same trick/hack to all binaries -
sorry for that.
I went ahead with reverting this ill-fated liaison with a libtool
convenience library - not so convenient after all... Exercise for the
reader: decide whether I meant libtool or its convenience libs.
As part of rebasing the dbus-server-reorganization-pohly branch I did a
few more things:
* include the AutoSyncManager syncDone fix (BMC #21888)
* fixed a compile error in the WebDAV backend due to the "Clean up
namespace pollution" patch
* moved D-Bus server sources into src/dbus/server and squashed
with the cleanup fixes for the original renaming to src/server
* removed accidental commit and its revert
Regarding the directory renaming we've had a misunderstanding. When I
said that the "dbus" should hold everything SyncEvolution D-Bus related,
that was meant to include the server.
The Notification* classes should also be moved there - not done yet, in
case that they need further work.
I'll probably have a look at compile times once all of this is in master
(= directly after 1.2 release). Pushing a lot of files into
sub-directories limits parallelism in "make -j", which is particularly
obvious on our shiny new nightly build machine (8 real cores). I also
saw on my laptop that compiling these *Register.cpp files multiple times
does consume noticeable time - perhaps there is a better solution after
all.
Tests are currently running. TestConnection.testCredentialsRight passed
already, so I am pushing my dbus-server-reorganization-pohly branch.
--
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.
4:02 p.m.
New subject: [SyncEvolution] dbus-server-reorganization progress
On Wed, Jul 27, 2011 at 2:52 PM, Patrick Ohly <patrick.ohly(a)intel.com> wrote:
On Mi, 2011-07-27 at 10:30 +0200, Chris Kühl wrote:
> On Wed, Jul 27, 2011 at 9:07 AM, Patrick Ohly <patrick.ohly(a)intel.com> wrote:
> > On Mi, 2011-07-27 at 01:45 +0200, Chris Kühl wrote:
> >> I've tried my dbus-server-reorg branch with this change as well as
> >> your dbus-server-reorg-pohly-libsynccommon branch but I still get the
> >> same results as before.
> >
> > Just to be sure, you mean the "no backends available in static
> > compilation" problem?
> >
> > The last commit fixed it for me. Anyway, the little benefit of compiling
> > the source files less often is not worth the hassle. I'll revert to
> > including the core source files in all executables.
> >
>
> This is the error I get for the TestConnection.testCredentialsRight for example,
> DBusException: org.syncevolution.InvalidCall: invalid value 'file' for
> property 'backend': 'not one of the valid values (virtual, calendar =
> events, addressbook = contacts, todo = tasks, memo = memos = notes)'
I forgot that we were talking about the syncevo-dbus-server. My patch
was only for the syncevolution binary. Because of its ugliness I must
have shied away from applying the same trick/hack to all binaries -
sorry for that.
Ah, which would have been obvious if I'd looked at the patch more closely. ;)
I went ahead with reverting this ill-fated liaison with a libtool
convenience library - not so convenient after all... Exercise for the
reader: decide whether I meant libtool or its convenience libs.
As part of rebasing the dbus-server-reorganization-pohly branch I did a
few more things:
* include the AutoSyncManager syncDone fix (BMC #21888)
* fixed a compile error in the WebDAV backend due to the "Clean up
namespace pollution" patch
* moved D-Bus server sources into src/dbus/server and squashed
with the cleanup fixes for the original renaming to src/server
* removed accidental commit and its revert
Regarding the directory renaming we've had a misunderstanding. When I
said that the "dbus" should hold everything SyncEvolution D-Bus related,
that was meant to include the server.
Ah, thanks for the clarification. However, it appears that you've gone
one level to deep. The source files in your branch are in
src/dbus/server/server/. Once I moved them up one level everything
worked as it should.
The Notification* classes should also be moved there - not done yet,
in
case that they need further work.
Yes, agreed. Looks out of place there.
I'll probably have a look at compile times once all of this is in
master
(= directly after 1.2 release). Pushing a lot of files into
sub-directories limits parallelism in "make -j", which is particularly
obvious on our shiny new nightly build machine (8 real cores). I also
saw on my laptop that compiling these *Register.cpp files multiple times
does consume noticeable time - perhaps there is a better solution after
all.
Tests are currently running. TestConnection.testCredentialsRight passed
already, so I am pushing my dbus-server-reorganization-pohly branch.
As I said, after moving the files that test passes on my end as well.
Cheers,
Chris
4:04 p.m.
New subject: [SyncEvolution] dbus-server-reorganization progress
On Mi, 2011-07-27 at 17:02 +0200, Chris Kühl wrote:
Ah, thanks for the clarification. However, it appears that you've
gone
one level to deep. The source files in your branch are in
src/dbus/server/server/. Once I moved them up one level everything
worked as it should.
Ups. I compiled, but not necessarily always from a clean directory.
Thanks for fixing 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.
12:47 p.m.
New subject: [SyncEvolution] dbus-server-reorganization progress
On Wed, Jul 27, 2011 at 2:52 PM, Patrick Ohly <patrick.ohly(a)intel.com> wrote:
<snip>
The Notification* classes should also be moved there - not done yet, in
case that they need further work.
Done. I've also removed the commit changing NULLs to 0s. These and a
fix for the files being in src/dbus/server/server are in my branch.
Cheers,
Chris
4039
days inactive
4040
days old
syncevolution@synevolution.org
3 comments
2 participants
participants (2)
-
Chris Kühl -
Patrick Ohly