personal opinion.....

Markus Feilner lists at feilner-it.net
Sun Oct 30 12:49:10 CET 2005


Am Sonntag, 30. Oktober 2005 11:45 schrieb Martin Konold:
> Am Sonntag, 30. Oktober 2005 10:58 schrieb Scott Newton:
>
> Hi,
>
> > > So what do you mean with "full imap support". Technically dIMAP is a
> > > superset of IMAP not the other way...
> >
> > My understanding of cachedimap (which may well be wrong) is that the mail
> > is stored in two places - on the server and on the client and they are
> > synced
>
> This "sync" term is unfortunately somewhat missleading but this subtle
> difference does not matter here.
>

Right. Cached Imap is something the Client does. 
I have several solutions running with a Cyrus Imap Server without Kolab, and 
offering shared (OFFLINE) ressources between the users on the IMAP Server.
Pretty much of Groupware is done by the client - Cyrus Imap offers pretty much 
possibilities already.
Kolab has a great webinterface for Administration (which cyrus has not 
natively), integration LDAP and IMAP really good.
And I love the idea of storing groupware data in IMAP mail folders - this 
offers great possibilities for team work, backup, and more.

However, there are several pitfalls: one is the package system.
And I still don't understand why kolab needs patched standard servers when 
other groupware solutions don't.

How can I help?

> > when the client is connected to the server. We were thinking of
> > something along the lines of a compressed linux image on a flash rom on
> > the terminal and the user keeping their basic configuration data on a
> > flash pen so it wouldn't matter which particular thin terminal they were
> > logged on to rather than having, say, an NFS share for the configuration
> > data. Under this scenario there wouldn't really be a client side to store
> > the mail data.
> >
> > As I said, at the moment it's just an idea - I'm just trying to figure
> > out if it is possible or not.
>
> It looks to me like you got the basic architectural points mixed up. As
> mentioned before your concerns don't apply to a thin client setup as the
> _only_ thing users need to carry around in a thin client setup is their
> token to do authentification. This token might be a password, a smartcard
> or a physical piece of the body (e.g. iris scan).
>
> The architecture you are describing is the old well understood
> client-server architecture not a thin client architecture.
>
> Last but not least the Kolab data fits perfectly on uptodate USB sticks...
>
> Regards,
> -- martin

-- 
Mit freundlichen Grüßen
Markus Feilner

--------------------------
Feilner IT Linux & GIS 
Linux Solutions, Training, Seminare und Workshops - auch Inhouse
Beraiterweg 4 93047 Regensburg Untere Hauptstr 2 85386 Eching
fon regensburg  +49 941 8107989 fon eching      +49 89 379 956 3 
fax +49 89 379 956 444  mobil + +49 170 3027092 
skype ID: mfeilner mail: mfeilner at feilner-it.net
SIP: +49 941 569 549 5 (sipgate)

-- 
Mit freundlichen Grüßen
Markus Feilner

--------------------------
Feilner IT Linux & GIS 
Linux Solutions, Training, Seminare und Workshops - auch Inhouse
Beraiterweg 4 93047 Regensburg Untere Hauptstr 2 85386 Eching
fon regensburg  +49 941 8107989 fon eching      +49 89 379 956 3 
fax +49 89 379 956 444  mobil + +49 170 3027092 
skype ID: mfeilner mail: mfeilner at feilner-it.net
SIP: +49 941 569 549 5 (sipgate)




More information about the users mailing list