About CalDAV, a new calendar storage specification ...
Helge Hess
helge.hess at opengroupware.org
Fri Jan 7 23:31:14 CET 2005
On Jan 5, 2005, at 11:03, Martin Konold wrote:
>> I know a XML storage format has been specified for Kolab2/Proko2.
> The cited URLs don't define a storage format but refer to iCalendar
> and define
> a transport / query mechanism.
iCalendar _is_ the storage format for CalDAV.
> My first impression is that they do not solve the problems which exist
> with
> iCalendar
There is the Calsify working group do work on that, its open, feel free
to join! CalDAV choose to clean up the existing approach instead of
throwing away all existing work and start something completely new.
> while not really addressing the scalability issues.
You just claim that there _are_ scalability issues. I claim that you
can write a client which is superior in scalability to Kolab - *if* you
know what you are doing and if you know what the protocol provides you.
HTTP solves scalability issues extremely well, this should be out of
question and is proven in practice.
> Basically CalDAV looks to me like an alternative database like
> approach without using
> SQL but using webdav instead.
Wow, three things mixed up in one sentence.
a) CalDAV is not a database, its a protocol
b) SQL is query language, not a protocol (the protocol is usually some
proprietary binary format)
c) WebDAV is not a query language, its a protocol (maybe you mix it up
with DASL)
>> And Sunbird (the mozilla calendar) has panned to implement it in its
>> roadmap
> Yes I am aware of this since many years.
You can't be aware of that, since CalDAV isn't even a year old.
> In general I still lack to understand the inherent advantage of webdav
> compared to imap for a secure, scalable and offline capable
> email/calendaring
> solution.
Since you ask, I answer ;-)
The only think which really matters to me is interoperability - Kolab
obviously has a very different focus - _which is just fine_. IMAP4
clients / servers - especially as used by Kolab - are VERY hard to get
interoperable. Its a huge mess to write an IMAP4 client which can
access Cyrus, Exchange, Domino, UW-IMAP, Courier at the same time. I
hope you won't question that. KMail, Evolution Mail and even OGo
webmail are *years of effort* to get the IMAP4/mail combination running
sufficiently well.
On the other side most of the 280 groupware solutions on Freshmeat are
based upon HTTP technology. This is why HTTP is a very good choice to
get people to implement a protocol to their storage. HTTP as the
transport has little to no interoperability issues - for none of those
solutions.
As another example take Kontact. It provides six connectors, one of
them - Kolab - is based on IMAP4. The remaining five are all based on
HTTP. Three of them (50%) are based on WebDAV.
CalDAV is intended to bring the 5 implementations together. Currently
there is no standard on how to use HTTP/WebDAV in the groupware context
which is why the above five are not interoperable. But they already
share a LOT of common base technology and adoption of CalDAV will be
way easier than adoption of IMAP4/Kolab-XML.
best regards,
Helge
--
http://docs.opengroupware.org/Members/helge/
OpenGroupware.org
More information about the users
mailing list