1:32 p.m.
I'm looking to setup server-to-server sync between Exchange and a
Caldav/Carddav server (Google probably, but memotoo for now). I'm looking
at https://syncevolution.org/wiki/ms-exchange-and-kde-synchronization, but
in the setup of the exchange link, I don't know which parts of "
target-config@Work *contacts*" are meaningful literals for syncevolution,
and which I can give names that are meaningful to me. As all the howto's
use 'target-config' so that might be a meaningful literal, but the howto at
the url also uses 'work@KDE'
1:59 p.m.
On Fri, 2014-03-21 at 14:32 +0100, Emiliano Heyns wrote:
I'm looking to setup server-to-server sync between Exchange and
a
Caldav/Carddav server (Google probably, but memotoo for now).
You really do like to try new things, don't you? ;-)
I'm looking at
https://syncevolution.org/wiki/ms-exchange-and-kde-synchronization,
but in the setup of the exchange link, I don't know which parts of
"target-config@Work contacts" are meaningful literals for
syncevolution, and which I can give names that are meaningful to me.
"target-config" is a special config which combines several properties
(like username and password) for sources created in the same context
(the part after @) and used on the target side of a local sync (the ones
with "syncURL=local://....").
"target-config" is a fixed string. The context name can be chosen
freely, for example "@google" for sources related to the same Google
account.
As all the howto's use 'target-config' so that might be
a meaningful
literal, but the howto at the url also uses 'work@KDE'
"work@KDE" is a sync configuration, the local side of the local sync in
this case. "@KDE" was used to make it explicit that the sources in that
context use KDE as local storage. The implicit "@default" context could
also have been used.
"work" is again a name that was chosen freely.
See the man page for a section on the terminology.
--
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.
2:26 p.m.
On Fri, Mar 21, 2014 at 2:59 PM, Patrick Ohly <patrick.ohly(a)intel.com>wrote:
On Fri, 2014-03-21 at 14:32 +0100, Emiliano Heyns wrote:
> I'm looking to setup server-to-server sync between Exchange and a
> Caldav/Carddav server (Google probably, but memotoo for now).
You really do like to try new things, don't you? ;-)
Guilty as charged. But I have a real use-case for this: my employer uses
Exchange, so I have all my work appointments there. I want my family to be
able to see both my personal and my work calendar. I can't give them access
to my Exchange account, but a sync between exchange and google would allow
them to see when I'm busy. One-way and two way sync would both work. I
could of course do this with an outlook plugin, but that requires me to
have outlook open almost always.
My current setup, I kid you not, is a VBox instance on my headless server
running Win7 running ICal4OL doing the syncing. But this setup fails
regularly for reasons entirely mystical, and even though I get some mild
amusement from proving Marc Andreessen right, it's not a setup I want to
keep.
> I'm looking at
> https://syncevolution.org/wiki/ms-exchange-and-kde-synchronization,
> but in the setup of the exchange link, I don't know which parts of
> "target-config@Work contacts" are meaningful literals for
> syncevolution, and which I can give names that are meaningful to me.
"target-config" is a special config which combines several properties
(like username and password) for sources created in the same context
(the part after @) and used on the target side of a local sync (the ones
with "syncURL=local://....").
"target-config" is a fixed string. The context name can be chosen
freely, for example "@google" for sources related to the same Google
account.
> As all the howto's use 'target-config' so that might be a meaningful
> literal, but the howto at the url also uses 'work@KDE'
"work@KDE" is a sync configuration, the local side of the local sync in
this case. "@KDE" was used to make it explicit that the sources in that
context use KDE as local storage. The implicit "@default" context could
also have been used.
"work" is again a name that was chosen freely.
See the man page for a section on the terminology.
OK, so if I get this correctly, I would have (in server to server sync) two
targets, one named (for example) target-config@exchange, and another
target-config@memotoo, and no local sync config (as I don't have KDE
running).
I see that the last parameter of the config command ("contacts" in the
howto) is supposed to be a 'source', but I can't tell between the howto and
the manpage what source it is supposed to refer to. The howto implies it's
connected to whether or not you use a template... but I haven't used a
template (so far). If I set up memotoo, I'll have to use the webdav
template, but
https://syncevolution.org/documentation/syncevolution-usage#caldav-and-ca...
the source parameter.
Also, when I set up two targets and no local source, how will
syncevolution --sync slow webdav
know that it's peer is exchange?
Thanks,
Emile
3:27 p.m.
On 21/03/14 14:26, Emiliano Heyns wrote:
On Fri, Mar 21, 2014 at 2:59 PM, Patrick Ohly
<patrick.ohly(a)intel.com
<mailto:patrick.ohly@intel.com>> wrote:
On Fri, 2014-03-21 at 14:32 +0100, Emiliano Heyns wrote:
> I'm looking to setup server-to-server sync between Exchange and a
> Caldav/Carddav server (Google probably, but memotoo for now).
You really do like to try new things, don't you? ;-)
Guilty as charged. But I have a real use-case for this: my employer uses
Exchange, so I have all my work appointments there. I want my family to
be able to see both my personal and my work calendar. I can't give them
access to my Exchange account, but a sync between exchange and google
would allow them to see when I'm busy. One-way and two way sync would
both work. I could of course do this with an outlook plugin, but that
requires me to have outlook open almost always.
Emile,
I wrote that HowTo and I apologize for the lack of clarity. My
understanding of the terminology is still pretty poor so I won't try to
answer our actual questions. But a couple of things you might find
useful from my experience...
I have a similar sort of problem -- my employer uses Exchange and I want
to have my appointments and my contacts available in other contexts and
on other devices. I chose not to try to do an actual server-to-server
sync but have set up an environment where I sync Exchange with a local
database and then sync that with various targets. In the example in the
HowTo, the local database is KDE but I have since replaced that with a
file backend. So, I sync Exchange with a set of files on my Linux system.
I then sync that set of files with an Owncloud server (caldav/carddav).
And I sync the Owncloud server with my phone (and other systems/devices
sync either with the files or with the Owncloud server).
Because I sync several folders (I have multiple contacts folders, for
example) and because I hack on the ActiveSync support, I have a script
which handles all the setup and the running of activesyncd and
syncevolution for me. I will send you that script off-list (I need to
remove some of my information, like my company's server address, before
I can send it). You may find some stuff in there to help you.
I see that the last parameter of the config command
("contacts" in the
howto) is supposed to be a 'source', but I can't tell between the howto
and the manpage what source it is supposed to refer to.
I find it easiest to think of the source as the "folder" -- it is the
place which stores the information about syncing one particular folder
(such as the Exchange folder name). The source name can be chosen as
you like -- but if you use a template, that template will automatically
set up some sources with particular names (personally I do not use
templates). But bear in mind that the source name is the default for
the uri field, which is the name used with the remote end in some
protocols (I *think* it is only important for SyncML).
It will be interesting to hear how you get on (how about writing up
another HowTo once you get it working?).
Graham
8:37 p.m.
On Fri, 2014-03-21 at 15:27 +0000, Graham Cobb wrote:
I wrote that HowTo and I apologize for the lack of clarity.
The key problem, at least initially, was different: Emiliano is running
the tool on a headless server. Identifying all env settings and daemons
which need to be started took a while - see
https://bugs.freedesktop.org/show_bug.cgi?id=76273
To summarize:
* One needs a D-Bus session.
* GNOME keyring is needed for activesyncd.
* The keyring must be password-less, because the keyring daemon
cannot show an unlock dialog.
* The Exchange password must be set via gconftool because
ssh-askpass also does not work.
* XDG_RUNTIME_DIR is needed on Ubuntu Saucy and later.
--
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.
8:41 p.m.
My howto will incorporate both aspects.
On Fri, Mar 21, 2014 at 9:37 PM, Patrick Ohly <patrick.ohly(a)intel.com>wrote:
On Fri, 2014-03-21 at 15:27 +0000, Graham Cobb wrote:
> I wrote that HowTo and I apologize for the lack of clarity.
The key problem, at least initially, was different: Emiliano is running
the tool on a headless server. Identifying all env settings and daemons
which need to be started took a while - see
https://bugs.freedesktop.org/show_bug.cgi?id=76273
To summarize:
* One needs a D-Bus session.
* GNOME keyring is needed for activesyncd.
* The keyring must be password-less, because the keyring daemon
cannot show an unlock dialog.
* The Exchange password must be set via gconftool because
ssh-askpass also does not work.
* XDG_RUNTIME_DIR is needed on Ubuntu Saucy and later.
--
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.
_______________________________________________
SyncEvolution mailing list
SyncEvolution(a)syncevolution.org
https://lists.syncevolution.org/mailman/listinfo/syncevolution
8:32 p.m.
On Fri, 2014-03-21 at 15:26 +0100, Emiliano Heyns wrote:
On Fri, Mar 21, 2014 at 2:59 PM, Patrick Ohly
<patrick.ohly(a)intel.com>
wrote:
On Fri, 2014-03-21 at 14:32 +0100, Emiliano Heyns wrote:
> I'm looking to setup server-to-server sync between Exchange
and a
> Caldav/Carddav server (Google probably, but memotoo for
now).
You really do like to try new things, don't you? ;-)
Guilty as charged. But I have a real use-case for this: my employer
uses Exchange, so I have all my work appointments there. I want my
family to be able to see both my personal and my work calendar. I
can't give them access to my Exchange account, but a sync between
exchange and google would allow them to see when I'm busy. One-way and
two way sync would both work. I could of course do this with an
outlook plugin, but that requires me to have outlook open almost
always.
As Graham said, it may be better to sync via a local storage. The
problem with direct syncing between Exchange and CalDAV/CardDAV is that
one side has to act as server (the one where the sync is started), and
that implies more frequent access to the data. It should work, it just
might be a bit slow.
Regarding CalDAV/CardDAV: I have not tested that with Memotoo. Memotoo
is regularly tested with SyncML.
Does Memotoo still convert all events into local time? That was the main
reason why I couldn't fully recommend Memotoo for calendar
synchronization. It's a lossy transformation that is particularly
painful in companies where recurring meetings get scheduled in different
time zones.
My current setup, I kid you not, is a VBox instance on my headless
server running Win7 running ICal4OL doing the syncing. But this setup
fails regularly for reasons entirely mystical, and even though I get
some mild amusement from proving Marc Andreessen right, it's not a
setup I want to keep.
Tip: while setting up syncing with SyncEvolution, permanently configure
loglevel=4 and increase maxlogdirs. That way you have full information
about previous syncs available and can analyze an issue even if you only
notice it later.
OK, so if I get this correctly, I would have (in server to server
sync) two targets, one named (for example) target-config@exchange, and
another target-config@memotoo, and no local sync config (as I don't
have KDE running).
No. "Local sync" got its name from syncing locally, on your own machine,
between two sources. Traditional SyncML syncing has most of the logic on
the server side, which is often remote and out of the control of the
user.
It does not matter whether the sources store data locally or remotely
because the SyncEvolution core doesn't even know that at all. There's no
flag saying "backend=caldav is a remote server".
The same applies to the terms "local" and "remote" in
"sync=refresh-from-local/remote": "local" is always the config which
is
used to trigger the sync, regardless of the actual backends.
In your case, you have a "sync config" (which is the one with
PeerIsClient=1 and syncURL=local://...) and a "target config" (the one
which is really called "target-config" in some context).
I see that the last parameter of the config command
("contacts" in the
howto) is supposed to be a 'source', but I can't tell between the
howto and the manpage what source it is supposed to refer to. The
howto implies it's connected to whether or not you use a template...
but I haven't used a template (so far).
"Sources" are the settings that define how to access data (backend,
database, potentially databaseUser/Password). This concept has many
names: folders in ActiveSync, collection in WebDAV, resources in
Akonadi, datastore, etc. - pick your poison.
Sources on the sync side are connected to sources on the remote side via
the "uri" parameter, which is the name of the remote source. All
SyncEvolution templates use well-known names (addressbook, calendar,
memo, todo) and match that to the corresponding data in specific peers.
If you don't use templates, you are free to choose whatever source name
you want.
If I set up memotoo, I'll have to use the webdav template, but
https://syncevolution.org/documentation/syncevolution-usage#caldav-and-ca... omits the
source parameter.
The Memotoo template uses SyncML.
Also, when I set up two targets and no local source, how will
syncevolution --sync slow webdav
know that it's peer is exchange?
Have a look at the "Synchronization beyond SyncML" section of the man
page (same as the README or the usage page [1]). That explains how the
two configs get connected.
[1] https://syncevolution.org/documentation/syncevolution-usage
--
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.
2978
days inactive
2978
days old
syncevolution@synevolution.org
6 comments
3 participants
participants (3)
-
Emiliano Heyns -
Graham Cobb -
Patrick Ohly